The Commissioner of Police of the Metropolis v The Information Commissioner & Anor

20. This Decision relates to an Appeal brought by the Metropolitan Police Service (‘MPS’) against the Information Commissioner’s (‘ICO’) decision notice (‘DN’) of 20 February 2023 which ordered disclosure of information requested under the Freedom of Information Act 2000 . The Second Respondent, George Greenwood (‘GG’) opposes the appeal on the basis that the DN was correctly decided. Factual Background to the appeal

21. The case has a long history. The factual and procedural chronology is detailed below – this is taken from the determination of the Upper Tribunal (before it was remitted back down to the First Tier Tribunal): ‘ Relevant factual background

5. Mr Greenwood is a journalist for The Times newspaper. In that capacity, he has written a number of articles about the Met Police’s handling of sexual offences. Following a reply by the Met Police to a prior request for information made to it by Mr Greenwood, on 15 August 2022 he made the following request for information to the Met Police: “Please provide a copy of the [Metropolitan Police Service’s] Pan London Serious Sexual Offence Problem Profile, as set out in your [previous reply].”

6. On 19 December 2022 the Met Police refused the request. It relied on sections 30(1)(a), 31(1)(a) and (b), 38(1) and 40(2) of the Freedom of Information Act 2000 (“FOIA”) in so doing. That decision was upheld on internal review.

7. Mr Greenwood then made a complaint to the Information Commissioner (“the ICO”) under section 50(1) of FOIA about the Metropolitan Police’s refusal of the above request. By a Decision Notice of 20 June 2023, the ICO rejected the Met Police’s reliance on sections 30, 31 and 38 of FOIA and required disclosure of the Serious Sexual Offences Problem Profile, save for personal data caught by section 40(2) of FOIA. In the course of his investigation of the complaint, the ICO had been told by the Met Police that it held one report falling within the scope of Mr Greenwood’s request. That report was 29 pages in length and contained: “tables, charts, maps, references, problem profiles, strategic assessments created as an intelligence product for the MPS. The report has a clear framework of analysis of information and intelligence for allowing a problem solving approach to law enforcement and crime prevention. The aim is to identify the issues and find ways to deal with the problem. The report covers key areas of concern for example identified and named venues, provides a detailed analysis regarding description of victim profiles and suspects. Names schools, areas for offences, venues for repeat locations, street names, hotel names, names of entertainment venues, names of mini cab companies, locations of sex workers, gaps and recommendations. The sensitive information contained within the report was not created or intended for public release as the information includes tactical elements as well as data and information relating to very specific areas, locations, and demographics which could also identify victims

8. The reasons why the ICO upheld Mr Greenwood’s complaint and found the Met Police had not dealt with his request in accordance with the requirements of Part I of FOIA were, in summary, because (a) the exemptions in sections 30(1)(a) and 31(1)(b) of FOIA are mutually exclusive and could not both cover the same information, and (b) the Met Commissioner, had relied on those exemptions and the exemption in section 38(1)(b) of FOIA in a blanket fashion and had not therefore demonstrated that any of the relevant exemptions were properly engaged.

9. Dissatisfied with this result, the Met Police appealed to the FTT under section 57 of FOIA. Under section 50(6) of FOIA, the making of this appeal had the effect of staying the steps the ICO had ordered to be taken under the Decision Notice. Mr Greenwood was subsequently made the second respondent to the FTT appeal.

10. In the course of the FTT proceedings, on 3 November 2023, the Met Police emailed Mr Greenwood and informed him that it had decided to “release the Problem Profile with only minor redactions to safeguard personal details, specific offence investigations and material that would cause prejudice to policing”. A copy of the redacted Problem Profile was attached to the email. The email went on to express the hope that the appeal could be dealt with by way of consent, and continued: “I am therefore refraining from serving today a statement explaining the application of the exemptions relating to the [redactions”], and the investigation correspondence referred to below.”

11. A few days later the ICO asked the Met Police to set out the exemptions it was relying on in relation to the redacted Problem Profile, which the Met Police did. On the face of it, the redactions to the Problem Profile went wider than those for personal data that the ICO had allowed for under section 40(2) of FOIA.

12. As result of the changed stance of the Met Police (to release the redacted Problem Profile as described in paragraph 10 above), on 18 December 2023 the ICO wrote to the Met Police and Mr Greenwood and said he had decided not to continue to defend his Decision Notice. He further said that he would be agreeable to the appeal being concluded by way of a consent order.

13. By way of a response, Mr Greenwood said he was not agreeable to the appeal being concluded by a consent order and he served a second witness in which he described what he considered to be the significance of the redactions made to the Problem Profile by the Met Police.

14. The Met Police then, on 2 January 2024, emailed the FTT as follows: “Dear Tribunal The Appellant no longer wishes to proceed with the Appeal of the Decision Notice, and the ICO no longer wishes to defend its Decision Notice. We are therefore giving notice under Rule 17(1) that the Appeal is withdrawn. We are aware that the permission of the Tribunal is required under Rule 17(2), and to that end the parties are currently working on a consent order to present to the Tribunal. In order to avoid the publicly funded bodies incurring unnecessary costs we do not intend taking any further steps in the proceedings, except in relation to the consent order.” The hearing of the appeal to the FTT was due to take place on 16 January 2024.

15. A further version of the redacted Problem Profile was disclosed by the Met Police to Mr Greenwood on 3 January 2024.

16. The Met Police’s email of 2 January 2024 caused the solicitors for Mr Greenwood to write to the FTT, on 4 January 2024, setting out that Mr Greenwood did not agree “to the Appellant’s withdrawal”. Mr Greenwood said that submissions on the withdrawal issue would be provided by him to the FTT the next day, and he asked that those submissions be taken into account by the FTT before reaching its decision “on whether to consent to the Appellant’s notice of withdrawal”.

17. On 5 January 2024 a registrar of the FTT gave case management directions on the appeal. These stressed the difference between withdrawing an appeal under rule 17, which would bring the appeal proceeding to an end and the ICO’s Decision Notice in place, and a consent order which would dispose of the appeal on terms agreed by the parties. The FTT registrar suggested that the stance of the parties suggested a consent order was unlikely to be agreed, and she asked the Met Police to confirm by 1pm on 9 January 2024 whether it (still) sought to withdraw the appeal.

18. In response to these directions, the ICO said he did not intend to be represented at the oral hearing on 16 January 2024. The full response reads as follows: “Dear Tribunal On the 3 November 2023 the Appellant (the public authority in this appeal) released the Problem Profile with redactions. The Commissioner has subsequently, and as explained to the parties, very carefully considered this appeal, the submissions and evidence, and raised queries with the Appellant on a closed basis which have been answered. The Commissioner has decided not to continue to defend his Decision Notice and explained this to the parties by email on 18 December 2023. The Appellant’s solicitor suggested the appeal be concluded by way of a Consent Order and the Commissioner has confirmed to the parties that he would be agreeable to a Consent Order with the exact terms to be agreed. The Commissioner notes the Second Respondent has indicated he does not agree to the appeal being concluded by way of a Consent Order and proposes the matter proceed to a substantive hearing. In a bid to reduce legal costs and to reduce the amount of judicial time incurred in considering this appeal, the Commissioner does not propose to file a skeleton argument or to be represented at the oral hearing on 16 January 2024. No discourtesy is intended, the Commissioner is merely attempting to preserve his and the Tribunal’s limited resources.”

19. The next day, 9 January 2024, the Met Police set out its response to the above directions, which reads: “Dear Sirs By email of 2 January the Appellant gave notice that the Appeal against the [Information Commissioner’s Decision Notice] was withdrawn. The Decision Notice stands. As stated below, in line with the overriding objective the Appellant and the First Respondent negotiated and reached an agreement regarding the subject matter of the appeal; there are no extant issues to resolve between them. Ideally the matter should be dealt with by way of a consent order, however this did not prove possible to achieve. The Second Respondent wishes to maintain the hearing listed for 16 January. This is a matter for him and the Tribunal. For the Appellant and First Respondent this hearing has no function. As public authorities, both are anxious to avoid expending further resources and therefore will take no further steps in the proceedings.”

20. The hearing of the appeal on 16 January 2024 was vacated and adjourned by the FTT of its own motion on 11 January 2024. The FTT in those adjournment directions stated that: “The Appellant’s request to withdraw the Appeal and the 2 nd Respondent’s opposition to this shall be considered and decided by the Tribunal on the 16 January 2024 on the papers.” The FTT reasoned that it was appropriate for the issue of withdrawal to be considered and concluded prior to the FTT and the parties being put to extra cost and expense in preparing for a full day hearing.

21. In a further email of 12 January 2024 to the FTT, in response to the FTT’s adjournment directions of 11 January 2024, the ICO set out the following: “I write with reference to paragraph 3 of the reasons in the Case Management Directions of Judge Heald dated 11 January 2024, the Commissioner had not agreed to withdrawal (and it does not appear this is a requirement of the Tribunal Rules in any event), but to a Consent Order with the terms to be agreed. The terms had not been agreed between the parties. It does not appear to the Commissioner that this appeal is amenable to a withdrawal. However it is noted the Tribunal will consider the Appellant’s application to withdraw its appeal and is considering this matter reflecting on the overriding objective, Rule 2. Following careful consideration of the submissions and evidence provided during the appeal, the Commissioner decided to change his position and agrees with the [Met Police] in relation to the application of the exemptions to the withheld information and no longer proposes to defend his Decision Notice. The Commissioner has suggested the appeal could be concluded by way of a Consent Order, with the exact terms of the Consent Order to be agreed, but acknowledged that the Second Respondent is not agreeable to this and wishes the matter to proceed to a substantive hearing. A Decision Notice can be substituted by two mechanisms, either a Consent Order or by a Tribunal Decision. The Commissioner would be agreeable to the Decision Notice being substituted with either a Consent Order or a Decision of the Tribunal which indicates that the appeal is allowed, and this Decision Notice be substituted with a new Decision that the Appellant was entitled to rely on the exemptions (setting out those exemptions) to refuse to provide the information requested and that the Commissioner does not require the Appellant to take any further steps in response to the request, with no order as to costs. If the Tribunal would be assisted by precise suggested wording from the Commissioner for the terms of a draft Consent Order then please let me know. It may be that following the Appellant’s email which was sent today at 17:11 the Second Respondent may decide to change his position, but that of course is a matter for him.” (the underlining is mine and has been added for emphasis)

22. I will return to this later, but it would seem that the view of the ICO which I have underlined in the above email was based on what the email later addresses about substituting the ICO’s Decision Notice of 20 June 2023; a Decision Notice which otherwise would, on the face of it, remain intact if the appeal was withdrawn. The FTT’s Consent to Withdraw Decision

23. The FTT’s decision consenting to the Met Police withdrawing its appeal was made on 16 January 2024….’.

22. The Second Respondent sought permission to appeal against that decision.

23. On 28 th February 2025, the Upper Tribunal allowed the second Respondent’s appeal. It set aside the First Tier Tribunal’s decision, consenting to the withdrawal, and remitted the case back to the FirstTier Tribunal to determine.

24. At this hearing, the IC confirmed in its skeleton argument (at paragraph 2) that it no longer supports the Decision: “The Decision was premised on the fact that the MPS’ original redactions lacked particularisation; however, the MPS has since withdrawn many of those redactions, and has provided (in the IC”s view) adequate explanations for which exemptions apply to the remaining redactions. Accordingly the steps required by the Decision – full disclosure of the Profile, save for the limited redactions under s.40 FOIA – no longer reflect the IC’s position”. Legal Framework General Right to Access to Information

25. Freedom of Information Act 2000 : ‘1. – General right of access to information held by public authorities. (1) Any person making a request for information to a public authority is entitled – (a) To be informed in writing by the public authority whether it holds information of the description specified in the request, and (b) If that is the case to have that information communicated to him…’.

2. Effect of the exemptions in Part II. … (2) In respect of any information which is exempt information by virtue of any provision of Part II, section 1(1)(b) does not apply if or to the extent that – (a) The information is exempt information by virtue of a provision conferring absolute exemption, or (b) In all the circumstances of the case, the public interest in maintaining the exemption outweighs the public interest in disclosing the information…’ Qualified Exemptions

26. Section 31(1) FOIA (the ‘Law Enforcement Exemption’) , states (insofar as relevant): ‘(1) Information which is not exempt information by virtue of section 30 is exempt information if its disclosure under this Act would, or would be likely to prejudice – (a) The prevention or detection of crime, (b) The apprehension or prosecution of offenders […]’

27. The ‘prejudice’ refers to prejudice which is a real possibility, or something which ‘may well’ result, even if the risk falls short of the higher test of ‘more probable than not’ Hogan and Oxford City Council v IC, IT, 17 October 2006 at [34] – [36]; John Connor Press Associates Ltd v IC, IT, 25 January 2006 at [15], following the approach of Munby J in R v Secretary of State for the Home Office ex parte Lord [2003] EWHC 2073 (Admin) .

28. The information need not relate to a specific investigation/prosecution.

29. The exemption can also apply to methods and procedures for investigating or preventing crime - not just information held for the purposes of investigation.

30. Particular information may appear innocuous but come within the exemption when considered with other (publicly available) information. The mosaic principle was adopted by the IT in in Hensley v IC and Chief Constable or Northamptonshire, IT, 10 April 2006 at [23].

31. Improbable harm, fanciful or improbable risks of harm to the relevant interest are to be disregarded.

32. If the information comes within the exemption, there is no duty to disclose if the public interest in maintaining the exemption outweighs the public interest in disclosing. The nature, degree and likelihood of prejudice to the prevention/detection of crime or the apprehension or prosecution of offenders needs to be balanced against and to outweigh the public interest in disclosure.

33. The public interest in maintaining the exemption will be stronger where disclosure would prejudice the prevention or detection of crime than where it is merely likely to have that effect.

34. Section 38(1) FOIA (the ‘Health and Safety Exemption’) states (insofar as relevant): ‘ (1) Information is exempt information if its disclosure under this Act would, or would be likely to – (a) Endanger the physical or mental health of any individual, or (b) Endanger the safety of any individual’

35. The exemption is a qualified exemption, so that whether or not there is a duty to disclose such exempt information will turn upon whether the public interest in maintaining the exemption outweighs the public interest in disclosing the information.

36. We note that Coppel (2023) states: ’32-011 Endangerment The exemption is unique amongst the exemptions in FOIA in its requirement of ‘endangering rather than ‘prejudicing’ a protected interest…some Tribunal decisions have held that there is no difference…it is suggested that the different terminology is deliberate and that it reflects the nature of the protected interest…it is suggested that ‘endanger’ will more readily be satisfied by any increased risk of the adverse consequence (harm to the physical or mental health or safety of any individual) but that it requires that the harm be more serious; whereas ‘prejudice’ places greater emphasis on the likelihood of the occurrence of that adverse consequence and less emphasis on the gravity of the adverse consequence’.

37. We agree with the interpretation adopted in Coppel. Nonetheless, for completeness we have applied both interpretations of the test in our consideration of the facts.

38. The likely effect of disclosure must be ‘real and not insignificant’ Hepple v IC and Durham CC, FTT, 26 February 2014 at [31]. The Tribunal must be able to identify a specific risk to the physical or mental health of at least one individual Hepple v IC and Durham CC, FTT, 26 February 2014 at [31].

39. Whether disclosure of particular information would/would be likely to endanger the health or safety of an individual may depend upon whether that individual has vulnerabilities or sensitivities particular to the individual. A public authority may be unaware of the particular vulnerabilities. The exemption does not permit a public authority to guess at such vulnerabilities or sensitives to evoke the exemption. Where the vulnerability or sensitivities are known to the public authority (because of other information that it holds) then the public authority will need to determine the application of the exemption on the basis of that knowledge.

40. This exemption is qualified – it only permits withholding the relevant information if the public interest in maintaining the exemption outweighs the public interest.

41. Hogan and Oxford City Council v ICO… (Appeal nos. EA/2005/2006 and EA/2005/0030) [§§28-35] set out a three-stage approach to the application of the prejudice test, which was cited with approval by the Court of Appeal in Zola [§22; §27] (emphases added): “First, there is a need to identify the applicable interest(s) within the relevant exemption… Second, the nature of the ‘prejudice’ being claimed must be considered. An evidential burden rests with the decision maker to be able to show that some causal relationship exists between the potential disclosure and the prejudice and that the prejudice is, as Lord Falconer of Thoroton has stated , ‘real, actual or of substance’. If the public authority is unable to discharge this burden satisfactorily, reliance on ‘prejudice’ should be rejected. There is therefore effectively a de minimis threshold which must be met. A third step for the decision-maker concerns the likelihood of occurrence of prejudice […] There are two possible limbs on which a prejudice-based exemption might be engaged. Firstly, the occurrence of prejudice to the specified interest is more probable than not , and secondly there is a real and significant risk of prejudice , even if it cannot be said that the occurrence of prejudice is more probable than not .”

42. Section 2(2)(b) requires that any prejudice which ‘would’ occur (i.e., ‘more probable than not’) or ‘would be likely’ to occur (i.e., ‘a real and significant risk of prejudice) under the law enforcement exemption, be balanced against the public interest in disclosing the information in “ all the circumstances of the case”.

43. The duty to communicate under s1(1)(a) is displaced by a qualified exemption under s2(2)(b) only if the public interest in maintaining the exemption outweighs the public interest in disclosure of the information sought; if the competing interests are equally balanced, the balance ought to fall in favour of communicating the information sought: Hogan [§56]. Absolute Exemption (i.e. not subject to the Public Interest Balancing Test)

44. Section 30(1)(a) FOIA (the ‘Criminal Investigations’ Exemption) states: ‘(1) Information held by a public authority is exempt information if it has at any time been held by the authority for the purposes of – (a) any investigation which the public authority has a duty to conduct with a view to it being ascertained – (i) whether a person should be charged with an offence, or (ii) whether a person charged with an offence is guilty of it’.

45. This is a class based exemption and there is no requirement for any prejudice to any investigation or proceedings to be shown for the exemption to apply.

46. The exercise of policing powers that are not, (at the time of the exercise) part of an investigation to ascertain whether a person should be charged with an offence or is guilty of an offence will not be within the first limb of the exception. However, if the public authority later holds it as part of an investigation to ascertain whether a person should be charged/is guilty, even if that was not the purpose for which it was acquired, it will be within the exemption.

47. Information once coming within the exemption is always potentially exempt and the potential exemption applies even if a decision was made not to prosecute/prosecution has been completed – however long ago the investigation or prosecution was completed.

48. There is no duty to disclose if and to the extent that the public interest in maintaining the exemption outweighs the public interest in disclosing the information.

49. The public interest in maintaining the exemption will be taken to include the factors identified in the White Paper, Your Right to Know: a. The need to avoid prejudicing effective law enforcement; b. The need to protect witnesses and informers; c. The need to maintain the independence of the judicial and prosecution processes; and d. The need to preserve the criminal court as the sole forum for determining guilt.

50. We note the following extract from Coppel (2023): ‘The passage of time, the completion of a prosecution, the release of similar information into a public forum and inconsequentiality of the information can all be expected to weaken the public interest in the maintenance of the exemption’.

51. This exemption does not cover information held within the prospect of some future criminal investigation or include all material that may be later used in any criminal investigation into a person not yet identified Williams v IC and Chief Constable of Kent Police [2021] UKUT 149 (AAC) ; [2022] 1 WLR 259 at [65-66].

52. Section 40(2) FOIA (the ‘Personal Information’ Exemption) states: ‘ (1) Any information to which a request for information relates is exempt information if it constitutes personal data of which the applicant is the data subject. (2) Any information to which a request for information relates is also exempt information is - (a) it constitutes personal data which does not fall within subsection (1) and (b) the first, second or third condition below is satisfied. (3A) The first condition is that the disclosure of the information to a member of the public otherwise than under this Act – (a) Would contravene any of the data protection principles, […]’

53. Section 3 of the Data Protection Act 2018 defines personal data, (in sofar as relevant) as: ‘(2) “Personal data” means any information relating to an identified or identifiable living individual. (3) “Identifiable living individual” means a living individual who can be identified, directly, or indirectly, in particular by reference to - (a) an identifier such as a name, an identification number, location data, or an online identifier, or (b) one or more factors specific to the physical, physiological, genetic, mental, economic, cultural or social identity of the individual’ The First Tier Tribunal’s role

54. The role of the First Tier Tribunal is set out in section 58 of FOIA: “If on an appeal under section 57 the Tribunal considers - (a) that the notice against which the appeal is brought is not in accordance with the law, or (b) to the extent that the notice involved an exercise of discretion by the Commissioner, that he ought to have exercised his discretion differently, the Tribunal shall allow the appeal or substitute such other notice as could have been served by the Commissioner, and in any other case the Tribunal shall dismiss the appeal. On such an appeal, the Tribunal may review any finding of fact on which the notice in question was based.”

55. The import of section 58 is that the right of appeal to the FtT involves a full merits consideration of whether, on the facts and the law, the public authority’s response to the FOIA Request is in accordance with Part I of FOIA ( Information Commissioner v Malnick and ACOBA [2018] UKUT 72 (AAC) ; [2018] AACR 29, at paragraphs [45]-[46] and [90]).

56. In accordance with the recent decision of the Upper Tribunal in Montague v Information Commissioner and DIT [2022] UKUT 104 (AAC) , at [86], “…the public authority is not to be judged on the balance of competing interests on how matters stand other than at the time of the decision on the request which it has been obliged by Part I of FOIA to make.” Preliminary Issues (i) Statement of Superintendent Davies

57. The Second Respondent submitted that the witness statement of Superintendent Davies should not be admitted into evidence. He addressed the reasons for this at some length in his skeleton argument – we do not set these out again here, other than a very brief summary. He submitted that the Tribunal should refuse to admit the witness statement into evidence under rule 15(2)(b)(iii) of the Tribunal’s rules (“ it would otherwise be unfair to admit the evidence” ). Alternatively, little weight, if any, should be afforded to the statement because (i) there are various alleged deficiencies in the evidence – as set out in their first skeleton argument; (ii) the Second Respondent is unable to cross-examine the witness, as such it would be unfair to admit it and inherently unreliable without being tested. Reliance was placed on the case of Al Rawi v The Security Service [2011] UKSC.

58. We heard oral submissions from all parties before making our decision. We noted the inability of the Second Respondent to test the evidence and attached weight to this (also acknowledging the reasons the Appellant was unable to attend, namely that he had retired). We noted that no real explanation was provided for why an alternative witness did not attend in his place. We indicated at the hearing, that we considered the evidence of Supt Davies was relevant and that it would be fair to admit the witness statement into evidence. We took into account that the witness was not cross-examined, nonetheless the relevant points that would have been taken in cross-examination were made in lengthy and detailed submissions (which greatly assisted us). We did not find it would be unfair to admit the evidence for that very reason. We concluded that the Second Respondent was able to (and did) respond effectively via submissions. He addressed both the weight to be attached to the statement and the veracity of the information contained therein.

59. We agreed with the submission made on behalf of the Appellant that the points of contest did not really go to difference of fact, but were more about opinions (and whether those opinions are justified) that flow into the exemptions. We agreed that this was “more a submissions type of case”. Detailed submissions were made. We took those submissions into account and paid significant regard to them. As a result we attached limited weight to the statement in various respects, as will be clear throughout this decision.

60. We note here the Second Respondent’s submission that much of Supt Davies’ evidence was speculation. We accepted the Appellant’s submission that the prejudice test is by its nature speculative. The Appellant submitted “ we don’t know whether there is going to be prejudice…It involves some speculation”. At paragraph 31 of their skeleton argument they submitted, referring to the ICO’s guidance (at paragraph 21) that it is not usually possible to provide concrete evidence that a prejudice would or would be likely to occur. We accept that imposing an expectation of concrete examples would create an unworkable FOIA regime. We did not expect concrete examples - nonetheless having heard the Second Respondent’s submissions we did find that Supt Davies, at times went beyond what was reasonable and therefore did speculate to such a degree that we attached limited weight to his evidence. (ii) Which exemptions did the Appellant rely on?

61. The Second Respondent submitted in its skeleton argument that the MPS cannot be permitted to rely on “ potential” s.30, s.38 or s.40 exemptions. As such, he addresses only s.31 in the skeleton argument.

62. The Appellant addressed the various exemptions in the skeleton argument and in oral submissions (we note that the Appellant made it clear in the grounds of appeal (Bundle Ai, pg 199) that sections 31 and 30 were relied upon ‘in the alternative’). The Second Respondent had the opportunity to respond to all the submissions – and did so – in oral submissions. We conclude that there was no unfairness or disadvantage to the Second Respondent from this. Discussion and Conclusions

63. For the purposes of determining this appeal, we have considered those documents contained within the 4 separate bundles provided to the Tribunal on the morning of the hearing. We were told that the documents had been served in good time, albeit they had not reached the Tribunal members, we accept this and make no criticism of the parties for the fact that we had not had sight of the bundles before the hearing, we note that we had read the majority of the papers in advance of the hearing – albeit in a different format): a. Bundle A (281 electronic pages); b. Bundle B (632 electronic pages); c. Bundle C: ’Materials Pre Withdrawal’ (60 electronic pages); d. Bundle C(i): ‘Withdrawal and Appeal’ (41 electronic pages); e. Closed Bundle (61 electronic pages); f. Appellant’s ‘Open Written Submissions for final merits appeal on 23 June 2025’ (24 electronic pages) g. First Respondent’s Skeleton Argument (6 electronic pages); h. Skeleton Argument for the Second Respondent for hearing on 23 rd June 2025 (33 electronic pages); i. Closed Material table of Redactions and Exemptions (9 electronic pages); j. Closed Withheld Profile showing MPS redactions (provided after the hearing).

64. All parties attended the hearing. No oral evidence was provided by any witness.

65. Following an Open Hearing, the Tribunal went into Closed, where the representatives for the Appellant and First Respondent made submissions. A Gist consisting of 6 paragraphs was provided to the Second Respondent and submissions were made by the Second Respondent in response.

66. As regards the gist and closed material, the Tribunal continued to consider its obligations as set out in Browning -v- Information Commissioner and [2014] EWCA civ 1050 Barrett v The Information Commissioner and Financial Ombudsman Service It maintained the decision to keep the documents closed. [2024] UKUT 107 (AAC) (20 April 2024).

67. Following the hearing, the Tribunal was provided with ‘the Closed Withheld Profile showing MPS redactions”. On further examination, it transpired that this document was not accurate – it had several errors. Nevertheless, it was of general assistance to the Tribunal. Discussion and Conclusions Section.38 of FOIA 2000

68. The Appellant submits that disclosure of certain redactions would be likely to ‘ either endanger the physical or mental health of individuals or endanger their safety’ (para 66 Grounds of Appeal). In the Closed Material table, the Appellant submitted that s.38 was applicable to the redactions at internal pages 8 and 12.

69. In its skeleton argument, the Appellant adds: ‘40. R1’s guidance titled ‘Section 38 – Health and safety’12 summarises relevant principles. The focus of s.38 is on non-personal information the disclosure of which may pose a risk. Endangerment requires more than mere prejudice. The endangerment of mental health refers to disclosure leading to psychological disorder or worsening mental illness: a threshold higher than causing upset or distress. Clinical psychiatric evidence is not always necessary, however.

41. Case law provides that it is not sufficient for the Tribunal to find that disclosure ‘might just conceivably lead to harm’. There must be a real and significant risk (albeit potentially short of being more likely than not) of the endangerment occurring (see Keane v (1) IC (2) Home Office (3) MPS [2016] UKUT 0461 (AAC) at §50 per UTJ Wikely)’.

70. No evidence of the anticipated harm was submitted to the Tribunal, other than a very brief reference in the witness statement of Superintendent Christopher Davies (for completeness, we add that we were not expecting clinical psychiatric evidence).

71. The Appellant accepts in the skeleton argument that endangerment requires more than mere prejudice, endangerment of mental health refers to disclosure leading to psychological disorder or worsening mental illness – a threshold higher than causing upset or distress.

72. We find that there is nothing to show that disclosure would cause anything more than upset or distress to the individuals in question (let alone leading to psychological disorder or worsening mental illness). Therefore, on the evidence before us, we concluded that disclosure would not be likely to endanger the physical or mental health or safety of any individual and therefore the exception is not applicable in the circumstances.

73. The Appellant further accepts in the skeleton argument that there must be a real and significant risk of the endangerment occurring. Again, based on the evidence before us we conclude that even if we are wrong and there is a risk of endangerment, it is too low to engage the legal threshold of s.38. The Appellant has not shown that there is a real and significant risk of the endangerment occurring. Nor has he demonstrated that it ‘ would be likely’.

74. When considering the particular vulnerabilities or sensitivities which would be relevant to these considerations. We note the submission that it is uncontroversial that there are serious mental health links from Serious Sexual Offences (“SSO”) to such victims – we took account of the article ‘ Understanding sexual violence and factors related to police outcomes’ in Bundle B at pages 602, which concludes that sexual violence has lasting effects on victims’ mental health. It was submitted that the argument was based on generality but one that is logical.

75. We took account of the submission that: “ The particular vulnerabilities of victims are also notable. They include minors, as well as adults, persons with mental health difficulties, learning difficulties, refugees, those from the LGBT+ community, persons with substance misuse issues and sex workers, among others. These features, alongside the sensitive and high-profile nature of the SSO, give rise to a relatively high degree of endangerment to affected individuals, such that s.38 is engaged ” (at paragraph 62 of the Appellant’s skeleton argument).

76. Even if it is correct that the individuals should be considered to be more ‘vulnerable’ than the ‘average’ person, we do not find the likely effect of disclosure to be ‘real and not insignificant’ (let alone reaching the higher test of increased risk of adverse consequences, which arguably is the test required through the use of the word ‘endanger’ (as set out at paragraph 19 above) – for completeness, we confirm that we did not apply this higher test).

77. Based on the evidence before us, we were not able to identify a specific risk to the physical or mental health of at least one individual Hepple v IC and Durham CC, FTT, 26 February 2014 at [31].

78. We did not find s.38 was applicable to any of the redactions. We address it no further in this judgement. Section 40(2) of FOIA 2000

79. The Second Respondent accepted in his response (Bundle Ai, pg 222, para 8): “the Commissioner correctly recorded that Mr Greenwood was content for personal information contained in the Profile which properly falls within the scope of s.40(2) FOIA to be withheld from disclosure. This remains the case. Mr Greenwood also recognises and accepts the prohibition in s.1 of the Sexual Offences (Amendment) Act 1992 (the ‘ 1992 Act’ ) on publishing any matter that is likely to lead members of the public to identify a complainant of a SSO, and the policy reasons which lie behind this prohibition, and he has absolutely no interest or intention of doing so’.

80. In oral submissions, the Second Respondent expanded on this, stating that there are two categories of s.40 data. First there is personal data which they accept ought not to be disclosed. The second category would be if the Appellant could show that one offence took place on a street and this information even when anonymised could be personal data because it could allow someone to be identified. The Second Respondent maintained that this was not plausible but if he was wrong on this, the next question would be whether it was unfair, unlawful or untransparent to disclose it. He maintained that the only identification of that person would be by someone who already knows about the offence and that would not be unfair or unlawful as it would be a tiny fraction.

81. The Appellant submits in the Closed Material Table, that s.40 is applicable to the redactions at: a. Internal page 1/Bundle B, pg 568 (both redactions); b. Internal page 7/Bundle B, pg 574 (first two redactions) the numbers were exceptionally low and the venue very specific; c. Internal page 8/Bundle B, pg 575 (first redaction); d. Internal page 12/Bundle B, pg 579 (all redactions); and e. Internal page 16, Bundle B, pg 583 (first three redactions).

82. Having examined the redacted sections, we agree that the following redactions are personal data and therefore are exempt under s.40: a. The first redaction on internal page 1; b. The second redaction on internal page 1. Both redactions are names and therefore s.40 is directly applicable. As such, the Second Respondent accepts that it ought not to be disclosed.

83. The rest are (arguably) the “second category” of s.40 data. We address them alongside the other exceptions below. The Four Categories of Redactions

84. The Appellant maintains that there are 4 categories of redactions, all of which rely on the s.31 exemptions. The first and third categories also allegedly engage ss.30, 38 and 40. We have addressed our findings re exemptions under s.38 above. We address s.30, s.31 and s.40 from this point onwards.

85. The 4 categories of redactions were: (i) Specific geographic locations, in the form of addresses and maps of reported crime such as streets, public spaces, schools, night clubs, hotels.

86. The Appellant states in its skeleton argument that: “44.1… It should not be controversial that disclosure of any location details that could result in mosaic/jigsaw identification…Pinpointing such venues and sites would risk the identification of such venues as potentially high harm locations, with little benefit in terms of alerting the public. The data in the information is by now relatively aged, again, limiting the benefits of disclosure. Perpetrators would be likely to be able to better identify victims or victim types and may be able to identify whether an attack perpetrated by them has been reported. This could inform future offending. (The mere fact of a report about a specific incident or relating to a particular location is highly valuable intelligence for offenders. That is because it may reveal the likelihood that a person has contacted the police about a particular offence)’ And “50. …The profile provides granular detail on specific areas and locations of concern. As such it identifies law enforcement methods and tactics as well as potential gaps in intelligence and recommendations… 51.2…The information contains granular detailed information on the venue of incidents. Disclosure may hinder the effectiveness of law enforcement in these areas or in relation to such incidents… “51.10. The MPS relies on the ‘mosaic effect’ in relation to prejudice to the law enforcement interests (as noted in R1’s guidance on section 31 at internal pages 9-10, §§21-23). Motivated and emboldened offenders are likely to use the information together with other information that is available to them in the public domain, to enhance their ability to avoid detection and/or to exploit actual or perceived law enforcement vulnerabilities. Furthermore, the mosaic effect would significantly enhance the ability of both offenders and members of local communities to identify VIWs who have contacted the police given the granular level of detail provided, in combination with other local knowledge or information in the public domain. Small datasets further enhance the risk of mosaic identification. The MPS’s precautionary approach is justified by the risks”.

87. The Second Appellant argues in its original skeleton argument that: “64…As to ‘specific geographic locations’, no real or actual prejudice can arise. The MPS already publishes on its website the same and/or substantively similar location information which enables the public to identify crime hotspots: see the evidence of Mr Greenwood at [ B/274/§10 ] and [ SB/557/§11, §13, §§16-18 ]. Accordingly, the same or similar information will already be in the public domain, often in granular and detailed form: see, for e.g., crime maps of the Paddington area [ SB/622 ] and the New Cross area [ SB/549 ] which detail exactly the number of crimes of the type ‘violence and sexual offences’ that have been reported in a particular month, pinpointed to such a granular level that nearby street names and/or local business and/or the nearest station or public space can be easily identified [ SB/622 ]. Similarly, the MPS already publish on its website names of streets, parks, and other specific locations in its local ‘crimes and priorities’ summaries, which often refer in granular detail to policing priorities in specific geographic locations: e.g., one of the MPS’ crime priorities issued on 3 August 2023 was “tackling ongoing ASB issues with street drinkers in and around Deptford High Street, Giffin Square and Douglas Way” [ SB/553 ].

65. The MPS has not properly explained why publication of locations of repeat offences on its website and elsewhere (at a granular level or detail including street names) would not prejudice law enforcement but that the locations of repeat offences contained in the Profile would, save that Supt. Davies in his evidence states that “while crime maps are available on the internet, in terms of detail and accuracy they cannot be compared to those created for internal police use” [ B/232/§7(i) ]. If it is the case that the maps contained in the Profile are for ‘internal police use’, and are different to those already published by the MPS and elsewhere, then the MPS have failed properly to explain how the maps are different (to the extent that by their publication the MPS suffers prejudice). The supposed detriment simply has not been made out.

66. Further, there is no consistency as to when ‘hotspot’ locations have and have not been disclosed in the redacted version of the Profile. The AW BCU summary page, for example, refers to Shepherd’s Bush Green (which is a green common and public space) as a ‘hotspot’ [ B/259 ], yet elsewhere in the Profile, names of other green public spaces such as forests/parks/heaths/commons have all been redacted [ B/247-248 ]. In some instances, the MPS have redacted location references while at the same time referring to those locations as recognised sites of problematic or criminal activity: e.g., “this has been a known problem area… for criminality, drug abuse and sex work” [ B/251 ] or “…is a well known location for street prostitution and drug addicted sex workers” [ B/261 ]. If these areas are ‘well known’ it is difficult to foresee why any harm or likelihood of prejudice arises such that redaction is justified. It appears that references to hotels, apps, and schools, have also been included in the ‘first’ category of redactions [ SB/566 ]. The names of hotels, schools, or apps cannot give rise to any harm or likelihood or prejudice such that redaction of such references could be justified. In relation to apps, it has been widely reported in the media that Uber is used by sexual offenders to provide access to victims, and is therefore already in the public domain, as is the fact that Uber recently introduced an emergency button in response to concerns relating to offences: [ SB/623 ]. In relation to schools and hotels, the MPS already publish hotspot maps (as discussed above at §64) from which the names of nearby schools and hotels can be identified: see, for e.g., ‘hotspot’ maps showing that 2 offences were reported in the immediate vicinity of Addey and Stanhope School [ SG/615 ]; 3 offences reported in the vicinity of Deptford Green School [ SB/616 ]; and 23 offences reported in the vicinity of Travelodge London Crystal Palace [ SB/618 ]. Therefore, notwithstanding that much of this information is already available in the public domain, it is not clear how disclosure to the public of the names of social media apps, schools and hotels (already disclosed as ‘budget hotels’ [ B/255 ]) could possibly give rise to any prejudice to the MPS’s ability to prevent or detect crime, or to prosecute or apprehend offenders ”.

88. We note from the outset that although there are crime maps available to the public, these are in many respects different to what is withheld in the report we are considering. As one example, looking at the crime map at page 550 of Bundle B, we note that it explicitly states “ About this crime map. This map uses data from local forces. The hotspots give only a loose idea of where crimes happened. Actual locations and details of crimes are kept anonymous. Please note that not all crimes that occurred can be shown on the map. 2295 incidents of Crime occurred in Metropolitan Police Service could not be mapped to a location and therefore are not on this map”. We accept the Appellant’s submission that there are several caveats in relation to the publicly available information – it is not as detailed nor as accurate as some of the maps in the report (we address these individually as we go through the redactions).

89. We agree - in part - with the Second Respondent, that there are occasions when the report states that an area is a well-known location for a particular type of offence, in those circumstances we conclude that there can be no prejudice in disclosing the redaction. We again address these as we go through the redactions.

90. The Second Respondent asked the Tribunal (in oral submissions) to consider the level of granularity in the closed session. We did so, in great depth – both during the hearing and in our deliberations following the hearing. We concluded that some redactions should be disclosed and some should remain redacted, we expand on the reasons for this below and in the closed schedule.

91. We concluded that the following redactions should be disclosed for the reasons provided below. We found no prejudice, for the reasons provided.

92. All the redactions on internal page 9 and 10, before the heading “Chemsex”. The Appellant relies on 31(1)(a) and/or (b), stating “ Specific parks are identified...Disclosure prejudices law enforcement and risks identification. The utility of this information under the PIT is limited by the age in the context of changing patterns” (as detailed in the closed material, table of exemptions). We considered these redactions should be disclosed as disclosure would not be likely to prejudice either the prevention or detection of crime or the apprehension of offenders. We found the evidence to be vague, simply referring to ‘parks’ - often very large areas, not specific places which would allow identification of individuals/specific offences – they were not accompanied by low numbers. It was not so detailed that it reveals new information. It was unclear to us how it would ‘prejudice law enforcement’ or ‘risk identification’. We were asked to find that the public interest was limited due to the age in the context of changing patterns. However, in light of the guidance provided in Montague, we did not attach any weight to this submission, as the relevant date for our considerations was the time the decision of the request was made. If we were wrong about prejudice, we found that the public interest favoured disclosing the information as it would allow the public to avoid the areas and we concluded that they would be more likely to report any incidents in those areas, knowing that other offences had taken place there.

93. All the redactions on internal page 11 and the top of page 12 (above the heading “sex workers”) were redacted on the basis of s.31(1)(a) and/or (b). The redactions on page 11 were withheld on the grounds that ‘ Low numbers redacted to avoid mosaic identification and prejudice to law enforcement’ (as detailed in the closed material, table of exemptions) . The redactions on page 12 were withheld on the basis that ‘ Names particular venues in the context of relatively low numbers of offences. There is a significant risk of mosaic identification and/or individual alarm or distress impacting on health & safety and victim confidence, prejudicing law enforcement interests and/or disclosing personal information unjustifiably’. In relation to internal page 12, the Appellant also relied upon ss.30, 38 and 40. We considered that this information should be disclosed as disclosure would not be likely to prejudice either the prevention or detection of crime or the apprehension of offenders. We accept that some of the numbers are low (especially at the beginning of page 11), however insufficient information is provided to identify individuals (e.g. there are references to low numbers in boroughs – but no further granular detail). We noted there was reference to offences having taken place in a park – however there are many parks in the boroughs referred to and no indication of which borough is being referred to. This would not lead to mosaic identification in our view, we did not find there was a real possibility of prejudice. We were not persuaded that it would prejudice law enforcement or the apprehension or prosecution of offenders. If the numbers were high, we could see how this could perhaps highlight police failings which could be exploited, but that was not the case here. We did not find s.38 was applicable for the reasons set out above at paragraphs 70-78. We did not agree that s.40 was applicable as we did not conclude that individuals were identifiable from this information. When considering s.30 considerations, we concluded that even if the exemption was engaged (noting that there is no requirement to demonstrate harm or prejudice that may result from disclosure), the lack of harm from disclosure was highly relevant to public interest considerations. We found that the public interest favoured disclosure in the circumstances. Keeping in mind that the exemption does not cover information held with the prospect of some future criminal investigation. Nor did it include all material that may be later used in a criminal investigation into a person not yet identified Williams v IC and Chief Constable of Kent Police [2021] UKUT 149 (AAC) ; [2022] 1 WLR 259 at [65-66].

94. The first redaction on internal page 12 (under the heading “Sex Workers”): the Appellant relied on s.31(1)(a) and/or (b) (and ss.30, 38 and 40). We disagreed with this analysis for the following reasons: i. The area referred to does not disclose a particular venue, nor a low number of offences (no numbers are actually disclosed – only the words “numerous offences” has already been disclosed). In these circumstances, we do not find that there would be a significant risk of mosaic identification; ii. Nor was there any evidence that this would prejudice law enforcement interests and/or personal information be unjustifiably disclosed - it was unclear to us how this could be the case. No information was being disclosed which could do this (for example nothing to suggest there would be undercover police in the area). The only disclosure is to say where it happened in the past; iii. We have already explained that we do not find s.38 is applicable. We concluded that disclosure would not be likely to prejudice the prevention of crime or the apprehension or prosecution of offenders, we therefore found s.31 was not applicable. There is nothing before us to show that s.30 was applicable (indeed the grounds for non-disclosure set out in the closed material table (set out above) do not explain how this is applicable,). Again we conclude that the public interest favours disclosure in light of the fact that we have concluded there is no risk of mosaic identification from disclosing this information. We concluded that s.40 was not applicable as an individual could not be identified from this redaction.

95. The first two redactions on internal page 13 (noting that the second redaction, spans 2 lines) and the map. The Appellant relies on s.31(1)(a) and/or (b). We concluded that the exemption was not engaged for the following reasons: i. In relation to the second redaction, we attached weight to the fact that the unredacted section of the same paragraph states that “ The W2 area from [redacted] also saw a high concentration of offences. This has been a known problem area…for criminality, drug abuse and sex work”. Where the area is a known problem area, there is no prejudice in disclosing the redacted sections. Further it is a broad area, not granular – we did not therefore conclude that there was a risk of mosaic identification. We have already disclosed the word contained under the first redaction above at paragraph above. We found there was no prejudice from disclosure. ii. In terms of the map, we considered that the detail contained therein is not granular enough to create harm (it is very difficult to understand exactly what the map is of – there are no road names, the resolution is very poor, the actual offences are not identified). It was unclear to us how or why this would prejudice law enforcement interests. The fact that unspecified offences relating to sex workers had taken place in a general area (which is all that is really shown by the maps, in our opinion), would not lead to mosaic identification, assist offenders or put victims off from reporting offences in our view. Again we found no prejudice from disclosure.

96. The following redactions on internal page 14 and 15: i. The first redaction (spanning two lines, on page 14); ii. The second redaction (page 14); iii. The fifth redaction (spanning two lines on page 14); iv. The maps on page 14; v. The maps on page 15. The Appellant relies on s.31(1)(a) and/or (b). We concluded that the exemption was not engaged. It is entirely unclear why ‘ the hotspot of Hoxton East & Shoreditch area ’ was disclosed on page 14 without engaging any of the arguments above, when other areas were not disclosed. There are no ‘venues’ disclosed (other than one on page 14 – third redaction which we have concluded should remain redacted below) – the rest are wards or districts. The maps do not seem to have a table, explaining what the different shading represents. There are no road names or venues identified. There are no numbers. No offences are specified. The maps seem to show a broad range of offences with some hot spots. It was not clear to us how the maps could give an advantage to any criminals nor how they could hinder ongoing investigations. Notably, the descriptions provided describe the maps. It was unclear why the wording can be disclosed but the map cannot. We concluded there was no prejudice from disclosure.

97. The following redactions on internal page 21: i. The third redaction (located in the first column); ii. The fourth redaction (located in the first column, spanning over two lines); iii. The sixth redaction (located in the second column).

98. The following redactions on internal pages 22: i. The second redaction (which is located in the second column); ii. The fourth redaction (which is located in the second column. The Appellant relied on s.31(1)(a) and/or (b) . We find that none of this was applicable to these redactions: the areas referred to are large areas, not venues and there are no small associated numbers, no venues are identified – we concluded disclosure could not lead to mosaic identification. Again we do not find that disclosure could lead to prejudice. In terms of the last redaction – a particular venue (i.e. with an address/area specified) was not identified. We did not find there would be any prejudice from disclosure.

99. The redactions on internal page 23 include: i. The first redaction; ii. The second redaction; iii. The third redaction; iv. The fourth redaction; Again, the Appellant relied on s.31(1)(a) and/or (b) and the same reasoning as above. We considered that these redactions should be disclosed because: i. Following the first redaction, the report states (unredacted) the area “ is a well known location for street prostitution and drug addicted sex workers”, if this information is well known, there would be no prejudice from disclosing the redacted words. ii. Again, the second redaction is followed by the words “ is a well known location for brothels”. The same reasoning applies – if this area is a well known area, we find there is no prejudice from disclosing the second and third redactions (they both refer to the same area); iii. Just before the second redaction, “ Deptford” within [redacted words]” has been identified. The smaller area (Deptford) has been identified – it is unclear why the larger area that it is part of should be redacted in those circumstances. The more granular information has been disclosed. We do not see any prejudice from disclosure. The area is disclosed in other paragraphs within this very section. iv. The third redaction refers to an ethnic identifier. However the sentence before states that the area “ is a well known location for brothels containing a high volume of [redacted] sex workers”. Again if it is well known specifically for sex workers of a particular ethnicity, there is no prejudice from disclosing this. We note that ethnic identifiers have been disclosed in other sections of the report – it was unclear why it could not be disclosed in this section, no submissions were made on this point.

100. The second and third redactions on internal page 24 (located at the top of the second column). The Appellant relied on s.31(1)(a) and/or (b) . We considered these redactions should be disclosed: the information redacted was very similar to the information already disclosed. The areas referred to were large and not granular. Indeed the only comment on them was that they were “ places of interest in regards to night time economy ”. There was no identification of crimes nor low numbers. We did not see how this could lead to mosaic identification. We address this in more detail in the closed schedule. We found there was no prejudice for this reason.

101. The redactions on page 26 include: i. The redactions in the last 2 bullet points in the first column; ii. The redactions in the first, second and fifth bullet points in the second column; iii. The first three words of the last redaction (last bullet point). It is unclear what exceptions or reasoning the Appellant relies on. The Closed Material table did not make reference to internal page 26. Nonetheless our role is to review the evidence and make our own decision. We found that the redactions above, all relate to large areas, they do not disclose ‘granular detail’ or specific areas with small numbers (indeed no numbers are included). They generally correlate with the map on the same page which has been disclosed. We did not find that it could lead to mosaic identification. They do not reveal granular details of policing gaps and methodologies in relation to specific areas (nor groups or profiles), each ‘area’ detailed above, is quite large. It was unclear to us how this could lead to any risks, as such we found no prejudice from disclosure. In relation to the last redaction, we concluded that the first three words should be disclosed as they have been disclosed in other parts of this decision.

102. The first redaction on page 28 (first column). Again, we found that this related to a large area and did not disclose ‘granular detail’ or refer to a road name accompanied by small numbers.

103. In contrast, we found that the following redactions should not be disclosed for the following reasons.

104. The first redaction and second redactions on internal page 7 (each spanning over two lines). These are very specific venues. It was submitted in the written submissions that “ There is a significant risk of mosaic identification and/or individual alarm or distress impacting on health & safety and victim confidence, prejudicing law enforcement interests”. The Appellant relied upon s.31(1)(a)(b)(c). We concluded ss.40 and 31(1) (a) and (b) were all applicable and it should remain redacted because the numbers are very low and the venue is very specific. It was submitted (in the closed hearing) that this could impact on the victims and that specific community, as information can travel far. We agreed that there was a real possibility of prejudice, keeping in mind, the specific community being referred to. We concluded that there was a causal link between disclosure and prejudice for the reasons already specified (i.e. disclosure of such low numbers and specific venues would be likely to lead to mosaic identification). The risk of prejudice is real and significant. We considered the public interest and found that maintaining the exemption outweighs the public interest in disclosure (noting that maintaining the exemption would be stronger if we had found that disclosure would prejudice the prevention or detection of crime than where it is merely likely to have that effect, which was what we had found). The Appellant acknowledged the public interest in disclosure at paragraph 56 of its skeleton argument, in particular that in a sensitive and high profile area, disclosure would likely build trust in the Appellant’s capabilities as well as hold it accountable for any perceived or potential gaps in intelligence and operational strategy. Mr Greenwood addressed, in detail, the public interest in the subject matter, in his witness statement, we noted in particular his evidence that, “ if the Met are failing properly to gather intelligence, or put sufficient capacity in place properly to address one of the most harmful forms of crime, there is a very clear public interest in this being put in the public domain for accountability purposes” (paragraph 19). We weighed this all in the balance. We further considered the submission that there is a public interest in the wider public being able to have a fully informed discussion on the subject matter, which may lead to other victims/witnesses/informants coming forward with relevant information and the suggestion that the public would be able to use the information to make their own decisions about safety. However, we concluded that this was outweighed by the fact that victims would likely feel more vulnerable and less willing to report crimes if they were identified or felt that they may be identified due to very low numbers. We considered factors set out in the White Paper, ‘Your Right to Know’, as set out above at paragraph 32. We agreed that disclosure of the information would likely result in victims feeling more vulnerable and threatened. We gave significant weight to the Appellant’s submission that “57.6.1 The severity of the impact of even one offender gaining a material advantage over MPS in its pursuit of the law enforcement interests is a weighty concern given that serious harm may follow. Even one such case would significantly outweigh the public interest in disclosure; 57.6.2 The severity of the impact of even one VIW (currently or in the future) being deterred from reporting an offence or supporting a prosecution is a weighty concern given their central role in securing justice. Even one such case would significantly outweigh the public interest in disclosure”. For completeness, we add here that we did not find that the likelihood of prejudice and the public interest in disclosure of the information were equally balanced - such that the balance should tip in favour of disclosure [ Hogan at paragraph 56]. We also concluded that s.40 was applicable as the granular detail in the information makes it possible for individual data subjects to be identified (or to identify themselves) due to the ‘ mosaic effect’ whereby a full picture is obtained by combining the disclosure with pre-existing information in the public domain, in particular the local community in this example. We found this could cause harm and distress to vulnerable individuals and inform offenders (indeed even the lack of reporting, provides them with information about whether there have been reports). We concluded that: - Disclosure would contravene the first data protection principle. It would be neither lawful, fair or transparent to disclose it, where the individuals have placed trust and confidence in the Appellant to keep their personal data secure and to protect them from risk of identification; - There is no basis upon which the disclosure of this special category data (data relating to a person’s sex life/data relating to criminal offences) may be permitted as being in accordance with the first data protection principle; - It is not in dispute that there is a legitimate interest in transparency of public authorities to inform public debate and to promote accountability. However, we agree that this is displaced by the rights of data subjects who remain at risk of identification and who may not be protected if their personal data were shared; - We do not find that disclosure to the world is necessary to advance that legitimate interest. We accept that there are other sources of information and scrutiny available; S.40(2) requires the requester to demonstrate legitimate interest for the disclosure. We find insufficient justification has been demonstrated. Data subjects who are victims of Serious Sexual Offending would have a reasonable expectation that information concerning their ordeals would not be disclosed to the world at large – reinforced by the effect of lifetime anonymity for victims under s.1 of the Sexual Offences Act 1992. We accept the submission that police must be able to reassure victims that their privacy will protected, as failure to do so may risk deterring victims from reporting offences; - Disclosure would likely result in unfairness to the individuals given the risk of identification and potential harm to their safety and harm to investigations.

105. The first redaction on internal page 8; the Appellant relies on s.31(a) and/or (b) and ss.30, 38 and 40. This is a very specific venue and once again involves very small numbers. It is very similar to those set out above. We rely on the same reasoning set out above to find that it should remain redacted under s.31(a) and (b), as well as s.40. We did not find s.38 applicable for the reasons stated in paragraphs 70-78. We saw no need to consider s.30 in the alternative.

106. The second redaction, under the heading “Sex Workers” on internal page 12 (spanning two lines): the Appellant relied on s.31(1)(a) and/or (b), s.30, s.38, and s.40. We did not find s.38 applicable for the reasons set out. However, we did find s.40 and s.31 applicable as the numbers are very low and the venue very specific, we agreed this could lead to mosaic identification. Again, we rely on the reasoning set out above to find there was a real and significant risk of prejudice and that there would be a causal link between disclosure and the prejudice. We rely on the same public interest considerations.

107. The third redaction on internal page 14 (following the words “A key repeat venue was…”): the Appellant relied on s.31(1)(a) and/or (b) on the grounds that “ Names of particular venues in the context of relatively low numbers of offences. There is a significant risk of mosaic identification, impairing law enforcement interests” . We found this was applicable as it identifies one very specific location. As such we found that both of the applicable interests in s.31 are relevant. The nature of the prejudice that may follow is a real possibility. There is a causal link between disclosure and prejudice and the likelihood of prejudice occurring is real and significant. Again, we rely on the same public interest considerations.

108. The redactions on internal page 16: the Appellant relies on s.31(1)(a) and/or (b) (and ss.30, 38 & 40). The submission made was that “ Names of particular venues in the context of relatively low numbers of offences. There is a significant risk of mosaic identification and/or individual alarm or distress impacting on health & safety and victim confidence, impairing law enforcement interests. R2 has accepted it is ‘perfectly reasonable’ for such details to be redacted”. In light of the exceptionally low numbers and specific addresses, we concluded that s.40 and s.31 applied (not s.38 for reasons set out at paragraphs 70-78). We agree that there was a real possibility of mosaic identification and resulting prejudice. As such there would be a causal link between disclosure and the prejudice. We find there was a real and significant risk of prejudice. We rely on the same public interest considerations. We found s.40 applicable again due to the high risk of identification. We did not need to consider s.30 in the alternative.

109. The first redaction (spanning 2 lines), the second redaction (spanning 2 lines), the third redaction (spanning 2 lines), the fourth redaction, the fifth redaction (spanning 2 lines) on internal page 17: the Appellant relies on s.31 (1)(a) and/or (b) on the basis that ‘ Specific apps and low numbers increases risk of identification and harm’. We agree that these should remain redacted: the numbers are very low, and, on some occasions, very specific addresses have been provided. We agree that this increases the risk of identification and harm. Where the Hotel addresses have not been provided, we concluded that these should still remain redacted because if an individual has committed an offence and the name of the hotel is not recorded, they will know that they have not been reported – it may give them confidence to reoffend. We found that both of the applicable interests in s.31 are relevant. There is a causal link between disclosure and prejudice (real, actual or of substance). And the likelihood of occurrence of prejudice is more probable than not and there is a real and significant risk of prejudice. Again we rely on the same public interest considerations

110. The fourth redaction on internal page 19: the Appellant relies on s.31 (1)(a) and/or (b) on the basis that ’Specific app/service increases risk of identification and harm’. We found that both of the applicable interests in s.31 are relevant. There is a causal link between disclosure and prejudice (real, actual or of substance). There is a causal link between disclosure: namely disclosing this number could lead to mosaic identification as it is a low number. We found the likelihood of occurrence of prejudice is more probable than not and there is a real and significant risk of prejudice. Even if it cannot be said that the occurrence of prejudice is more probable than not. Again we rely on the same public interest considerations as set out above.

111. The fifth redaction on internal page 21 (located in the second column): the Appellant relies on s.31 (1)(a) and/or (b) on the basis that “ Particular venues…and apps…includes highly granular accounts of particular and repeat offences, reports and patterns”. In light of the very specific venues detailed in the redacted words, we concluded that s.31 did apply. We found that both of the applicable interests in s.31 are relevant. There is a causal link between disclosure and prejudice (real, actual or of substance). The likelihood of occurrence of prejudice to the specified interest is more probable than not and secondly there is a real and significant risk of prejudice, even if it cannot be said that the occurrence of prejudice is more probable than not. Again we rely on the same public interest considerations.

112. The redactions on internal pages 22 include: i. Third redaction (located in the second column); ii. Fifth redaction (located in the second column, spanning 2 lines); iii. Sixth redaction (located in the second column, spanning 2 lines); iv. Map; v. Last redaction (located in the second column). The Appellant relies on s.31(1)(a) and/or (b) on the basis that “ Particular venues…and apps and relatively high resolution maps enabling broad pinpointing of locations. Includes highly granular accounts of particular and repeat offences, reports and patterns (eg 24). There is also tradescraft information about victim selection (eg 24)” We found that both of the applicable interests in s.31 are relevant. There is a causal link between disclosure and prejudice (real, actual or of substance) because the specific hostel names, road names, hospitals names, together with the map, considered together with the information disclosed under ‘Recommendations’ could lead to offenders avoiding these areas in order to evade detection. We found the likelihood of occurrence of prejudice to the specified interest is more probable than not. Again we rely on the same public interest considerations.

113. The redactions on internal page 23 and page 24 include detailed Maps. The Appellant relied on s.31(1)(a) and/or (b) on the basis that ” Particular venues (incl. schools / roads / clubs etc) & apps and relatively high resolution maps enabling broad pinpointing of locations. Includes highly granular accounts of particular and repeat offences, reports and patterns (eg. 24). There is also tradecraft information here about victim selection (eg. 24)” . We agree that the maps are relatively high resolution maps enabling broad pinpointing of locations – as well as direct identification of the type of offences through the table (in relation to the map on page 23). We accept that this is distinct from information in the public domain. As such we concluded that the risk of mosaic identification was significant and it should remain redacted under s.31. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime; and (ii) apprehension and prosecution of offenders. We further concluded that there was a causal link between disclosure and prejudice due to the risk of mosaic identification – this passes the de minimis threshold which we found met. We found that the likelihood of prejudice is likely such that there may very well be prejudice to the claimed interests even if the risk falls short of being more probable than not. We considered the public interest test but again considered that maintaining the exemption outweighs the public interest in disclosure.

114. All the redactions on internal page 25: the Appellant relied upon s.31(1)(a) and/or (b). The grounds for non-disclosure are “ Granular details of policing gaps and methodologies in relation to specific areas, groups or profiles” (in relation to the redactions at internal pages 25, 27-29). We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime and (ii) apprehension and prosecution of offenders. We further concluded that there was a causal link between disclosure and prejudice: they list very specific offences and detailed addresses, which when considered alongside the detailed maps, we accept could lead to mosaic identification or lead to offenders knowing their offences have been reported/victims identifying themselves – this passes the de minimis threshold which we found met. We found that the likelihood of prejudice is likely such that there may very well be prejudice to the claimed interests even if the risk falls short of being more probable than not. We considered the public interest test but again considered that maintaining the exemption outweighs the public interest in disclosure. Under the heading ‘Recommendations’ on the same page there is one further redaction. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime; and (ii) apprehension and prosecution of offenders. We further concluded that there was a causal link between disclosure and prejudice: the redaction details the specific area where the police will be increasing their presence. We accept that this discloses police tactics such that offenders may choose to avoid those areas in order to avoid detection and prosecution – this passes the de minimis threshold which we found met. We found that the likelihood of prejudice is likely such that there may very well be prejudice to the claimed interests even if the risk falls short of being more probable than not. We considered the public interest test but again considered that maintaining the exemption outweighs the public interest in disclosure.

115. In relation to “apps”, we found that much of the information relating to apps is already in the public domain or has already been disclosed in other reports. For this reason, we concluded that there was no prejudice to the Appellant’s ability to prevent or detect crime, nor to prosecute or offend offenders in disclosure of the redacted words. As such, we direct that the following redactions are disclosed : - Internal page 10 (under the heading “Chemsex” redactions 1, 2 and 3); - Internal page 11 (redaction 9 (or paragraph 4, redaction 2). We noted that the numbers were low however concluded this would not lead to mosaic identification as there was such little detail about the offences (e.g. unlike those redactions that we considered above, it was not disclosure of a particular type of offence at a particular address); - Internal page 18 (under the heading “Media or Online” redactions 1, 2, 3, (spanning over 2 lines), 5, 6, 7, 8). We again noted that some of the numbers were low however concluded this would not lead to mosaic identification as there was such little detail about the offences; - Internal page 19 (first, second and third redactions) and 20 (all redactions); - Internal page 21 (first two redactions, first column); - Internal page 22 (first redaction, first column); - Internal page 26 (second from last redaction).

116. In relation to hotels, we disagree that the two first redactions on internal page 17 should be disclosed. We have already addressed this above. (ii) Victim vulnerabilities, identifying particular characteristics that make individuals more likely than the general population to be the victim of sexual offending

117. The Appellant submits in its skeleton argument: ’44.2 MPS relied on the likelihood that offenders would not necessarily be aware that certain persons with particular vulnerabilities would be more likely to be at risk of SSO. Again, this could inform future offending. Any benefit in terms of increased public awareness are outweighed by the identification of vulnerable types of victims because they would be put at greater risk from motivated offenders’.

118. The Second Appellant argues in its original skeleton argument that: ‘68. As to ‘victim vulnerabilities’, no real or actual prejudice can arise. Supt. Davies states in his evidence that “identifying particular characteristics… make individuals more likely than the general population to be the victim of sexual offending” [ B/233/§7(ii) ]. However, ‘particular vulnerabilities’ are widely discussed in the public domain, including in publicly available academic literature: see the evidence of Mr Greenwood [ SB/557/§12 ]. The MPS does not address how other publication of vulnerabilities, and the public discussion thereof, would not prejudice law enforcement but that publication of the particular vulnerabilities contained in the Profile would so prejudice the MPS’s ability to prevent and detect crime and/or to apprehend or prosecute offenders.

69. The MPS’ approach to ‘vulnerabilities’ is also inconsistent. For example, in the thematic summary of ‘stranger offending’, the MPS states that “victim vulnerabilities through alcohol were highlighted as a common theme” [ B/251 ], whereas elsewhere, including on that same page, specific vulnerabilities have been redacted [see, for e.g., the final sentence of B/251 ]. Mr Greenwood refers in his evidence to other inconsistencies as to the MPS’ approach to vulnerabilities, including references to alcohol and drug use, gang membership, and the LGBT+ community [ SB/557/§12 ]’.

119. The Second Respondent submitted in oral submissions that what they understand Supt Davies to be saying is that if the public are told that certain categories of individuals have been victims more frequently than others, that would make them targets. It was submitted that this was speculative assertion, particularly when there are huge swathes of discussion about crime affecting certain groups more than others for example those with special needs or those who drink alcohol are at a higher risk. This is discussed regularly and is a valuable thing. They focused on the fact that various ‘vulnerabilities’ had been disclosed. It was unclear why others could not be. The Appellant responded that there was no inconsistency between some vulnerabilities being disclosed and others not being disclosed, it was about the context in which those characteristics were being referred to. Under each profile, there is a breakdown into years and the further breakdowns in relation to issues affecting those groups. It was at this point that other types of vulnerabilities were specified and there was a risk of prejudice and mosaic identification. This was what was behind the protective approach that had been adopted. Although it was a fair point to make in the abstract (that certain groups of people in society or categories of people are likely to be at greater risk of sexual offences), he did not draw on the broader societal context but on the specific context of the document. In this very specific context, it referred not just to reports of crime but data which has been identified as offences. It related to particular areas or time periods or even more granular geosocial factors. This was the reason the Appellant applied the exemptions.

120. We agreed with the Second Respondent that particular vulnerabilities are widely discussed in the public domain, including in academic literature. We also agree that numerous ‘vulnerabilities’ are identified in either other sections of the report or in previous reports released by the police. We did not agree with the Appellant that the context in which those characteristics were being referred to justified redaction of that information, although we continued to consider whether mosaic identification was a realistic possibility throughout.

121. We direct that the following redactions are released :

122. The last redaction on internal page 7 (spanning over 2 lines): the Appellant relied on s.31(1)(a) and or (b), on the grounds that “…There is a significant risk of mosaic identification and/or individual alarm or distress impacting on health and safety and victim confidence, prejudicing law enforcement interests. Low numbers exacerbates risk...” . It was submitted that the redaction identified 2 potentially vulnerable victim types by specifying the ‘ most prominent vulnerabilities’. We considered this should be disclosed as disclosure would not be likely to prejudice either the prevention or detection of crime or the apprehension of offenders. We did not find that disclosure would put those individuals at risk of this because offenders ‘ would not be aware that certain person with particular vulnerabilities would be more likely to be at risk of SSO’ when this information is already in the public domain. Nor did we find that the very specific context of the disclosure would mean there was prejudice or significant risk of mosaic identification. The numbers were not low and would not therefore enable this. We find that the vulnerabilities have already been identified and disclosed by other forces, as such there is no prejudice. We note that there was no qualitative difference between the source of the publicly available information and the information contained in this report.

123. The second redaction on internal page 8: the Appellant relied on s.31(1)(a) and/or (b) (& ss.30, 38 & 40). On the basis that “… There is a significant risk of mosaic identification and/or individual alarm or distress impacting on health & safety and victim confidence, prejudicing law enforcement interests and/or disclosing personal information unjustifiably. Low numbers exacerbate risk…”. Again we considered this should be disclosed as disclosure would be unlikely to prejudice either prevention or detection of crime nor the apprehension of offenders. We concluded that the fact that individuals with this vulnerability are targeted has already been disclosed by different police forces – the source of the publicly available information is an important factor, there was no qualitative difference between the two. We also found there was no risk of mosaic identification as the numbers were not low. As such, there was no prejudice from disclosure. We have already explained why we found that s.38 was not applicable. We did not find s.40 was applicable as the numbers were too high for this. When considering s.30, we concluded for the reasons set out above - in particular the existence of information in the public domain - significantly reduced the importance of releasing similar information. In conclusion we found that the public interest weighed heavily in favour of disclosing the redaction.

124. The last redaction at the bottom of internal page 10 (spanning 3 lines): the Appellant relied on s.31(1)(a) and/or (b). We concluded that this did not disclose anything new that hasn’t already been disclosed by the unredacted sentence preceding it. As such, there was no prejudice to disclosure. There is again no risk of mosaic identification as no numbers are provided, the reference to ‘victims’ makes it clear it is reference to more than one individual.

125. The last redaction at the bottom of internal page 13: the Appellant relied on s.31(1)(a) and/or (b) on the grounds that ‘ Names of particular venues in the context of relatively low numbers of offences. There is a significant risk of mosaic identification prejudicing law enforcement interests…Also references to particular vulnerability decreasing safety and impairing law enforcement’ . Again, we found these vulnerabilities had already been disclosed in other parts of the report or in other reports and therefore there is no prejudice from disclosure. We found there was no risk of mosaic identification as no numbers are provided.

126. The redacted table at the bottom of internal page 17 and the first words of the redaction above it: the Appellant relied on s.31(1)(a) and/or (b), on the following basis “ Specific venues (as above). Particular vulnerabilities/characteristics and evidential problems relating to repeat offending” . We saw no reason not to disclose this information. Some of the same information has been disclosed in the paragraph above. Other vulnerabilities have been disclosed throughout the report or in other documents, so there was no prejudice from disclosing them again. We accept that some other “victim characteristics” are listed here, which are not listed elsewhere, however we did not agree that (and could see no reliable evidence to show that) persons with particular vulnerabilities would be more likely to be at risk of SSO if this information was disclosed or that they would be at greater risk from motivated offenders (we also noted that we could not see how some of the headings could be described as “characteristics”). We could not find a causal link between disclosure and the prejudice/harm. We noted that the numbers for some categories were quite low but considered that this would not lead to a risk of mosaic identification as it is general information, it is not linked with a particular area (for example), especially as there was nothing to indicate that the same individual could not fall into several of these categories.

127. The redacted table at the top of page 18: the Appellant relies on s.31(a) and/or (b) on the basis that ‘ Specific apps and low numbers increases risk of identification and harm’. There is no reference to why the table identifying Offender Characteristics should not be disclosed. We accept that some of the numbers are low, however it was unclear to us how this would lead to mosaic identification of a victim, it relates to the offenders. Further it does not link the offender characteristics to an address or something that would permit them to even identify themselves. We could not see a causal link between the disclosure and prejudice. We considered that it may help offenders become aware of what ‘characteristics’ the police monitor. However, we could not see how this ‘intelligence’ would prejudice the prevention or detection of crimes nor apprehension or prosecution of offenders. It is unclear how or why offenders would change their behaviour in light of this information. Even if we were wrong about this, we found that the public interest weighed in favour of disclosure because it would actually assist the public to know what types of ‘characteristics’ (or arguably, what types of tactics) are deployed when committing these offences, so that they can be alert to them and try to avoid them as much as possible.

128. The redactions on internal page 18, under the heading “Media or Online”, in particular redaction 4 (spanning over 2 lines), redaction 9 (spanning over 2 lines), and redaction 10. Again, the Appellant relied upon s.31(1)(a) and/or (b) and the same reasoning as above. We found that these vulnerabilities had already been disclosed in other parts of the report or in other reports and therefore no prejudice arose from disclosure.

129. The fourth redactions, the fifth redaction and the sixth redaction on internal page 24 (all located in the second column): the Appellant relied on s.31(1)(a) and/or (b) on the basis that: ‘ Particular venues (incl. schools / roads / clubs etc) & apps and relatively high resolution maps enabling broad pinpointing of locations. Includes highly granular accounts of particular and repeat offences, reports and patterns (eg. 24). There is also tradecraft information here about victim selection (eg. 24)’ (these were the grounds for disclosure relating to internal pages 21-24) . We considered these should be disclosed: i. In relation to the fourth redaction, we concluded this should be disclosed as the redacted words are simply examples of what has already been discussed in the rest of the report and in particular in the paragraph that followed. As such there was no prejudice from disclosure; ii. In relation to the fifth redaction, we concluded that the first 6 words should be disclosed as there was nothing about these words that would be likely to prejudice either the prevention or detection of crime or the apprehension of offenders – it is does not identify any people or category of people; iii. In relation to the sixth redaction, we concluded this should be disclosed as the redacted words are simply examples of what has already been discussed in the previous sentence. Again, there would be no prejudice from disclosure.

130. The first three words of the last redaction on page 26. It is unclear, based on the Closed Material table, why the Appellant maintains this should remain redacted. We concluded that this information has been provided in other sections of the report which we have directed should be disclosed (namely the table at internal page 18). In those circumstances, there is no prejudice from disclosing it here.

131. In contrast, we found that the following redactions should not be disclosed :

132. The fourth redaction on internal page 14 (spanning 3 lines) . The Appellant relies on s.31(1)(a) and/or (b) on the following grounds “ Names of particular venues in the context of low numbers of offences. There is a significant risk of mosaic identification, impairing law enforcement interests….Also references to particular vulnerability and lower reporting rate that is not otherwise publicly acknowledged” . We accepted that this should remain redacted. It is information that discloses targeting of particular vulnerable groups and more importantly a lower rate of reporting for particular vulnerable groups. We concluded this could encourage offenders to target those particular groups, feeling confident that they would be less likely to be reported. As such we found that both of the applicable interests in s.31 are relevant. We concluded that disclosure would be likely to prejudice the prevention and detection of crime and the apprehension and prosecution of offenders. We found that there was a causal link between the disclosure and prejudice because the redacted words expose the benefit of targeting a particular vulnerable group, as such it gives a ‘strategy of offending’. In those circumstances we found there was a real and significant risk of prejudice. Again we found that the public interest in disclosure was outweighed by the exemption in particular because even one offender gaining a material advantage would outweigh the public interest.

133. The second to last redaction (spanning 3 lines) on internal page 17 – apart from the first 3 words (below the heading “Victim Characteristics” and above the table): the Appellant relies on s.31 (1)(a) and/or (b) on the basis that ‘ Specific apps and low numbers increases risk of identification and harm’. We found that both of the applicable interests in s.31 are relevant. We concluded that disclosure would be likely to prejudice the prevention and detection of crime and the apprehension and prosecution of offenders. We found that there was a causal link between the disclosure and prejudice because the redacted words expose a tactical advantage of targeting particular groups, and as such it gives a ‘strategy of offending’. In those circumstances we found there was a real risk of prejudice. Again we found that the public interest in disclosure was outweighed by the exemption in particular because even one offender gaining a material advantage would outweigh the public interest.

134. In relation to the fifth redaction on page 24, we concluded that the 5 last words should remain redacted. The Appellant relied upon s.31(1)(a) and/or (b) on the basis that “ Particular venues (inc. schools/roads/clubs etc) & apps and relatively high resolution maps enabling broad pinpointing. Includes highly granular accounts of particular and repeat offences, reports and patterns (eg 24). There is also tradecraft information here about victim selection (eg 24)” (this was in relation to the redactions on internal pages 21-24) . We agreed that this redaction seemed to be describing a highly granular account of a particular offence. We concluded that it could lead to mosaic identification. An offender may become aware that their offence has been reported and thereby change their behaviour (i.e. by targeting individuals different to those described) and that the disclosure would lead to that prejudice. We found that both of the applicable interests in s.31 are relevant and that disclosure would be likely to prejudice the prevention and detection of crime and the apprehension and prosecution of offenders. We found that there was a causal link between the disclosure and prejudice and a real risk of prejudice. Again, we found that the public interest in disclosure was outweighed by the exemption in particular because even one offender gaining a material advantage would outweigh the public interest. (iii) Low Numbers from small datasets, disclosure of which would greatly enhance the risk of particular individuals being identified

135. The Appellant states in its skeleton argument that: “44.3… Where the dataset is small, disclosing actual figures could enhance the risk of victim identification. Instead, MPS has disclosed the relevant percentages”.

136. The Second Respondent argues in its original skeleton argument that: “no real or actual prejudice can arise. Supt. Davies states in his evidence that disclosure of low numbers “would greatly enhance the risk of particular individuals being identified” [ B/233/§7(iii) ]. It is not clear how that risk materialises in practice. Small datasets do not automatically lead to a real risk that an individual will be identifiable: see the evidence of Mr Greenwood at [ SB/558/§15 ]. Further, it is not clear what criteria have been used to define numbers as ‘low’ so it is difficult to ascertain whether this approach has been consistently applied: for example, references to 9 cases of sex workers being ‘followed on the street’ and 9 cases” and 9 cases of sex workers being ‘beaten/punched/hit’ have not been redacted [ B/250 ].

137. The Second Respondent argued in submissions that paragraph 29 of Supt Davies’ statement made no sense. He was arguing that disclosing information that there has been a crime would allow someone who knows there has been a crime to know who the victim of the crime was. However, the person who committed the crime already knows who the victim is. It doesn’t tell anyone who doesn’t already know anything that could identify the individual.

138. We did not agree with the Second Respondent’s submissions in relation to this. We concluded that the risk could materialise in practice (albeit we agree that it would not necessarily be in every case). If the information contained is granular enough, we concluded that it would enhance the ability of both offenders and members of the local community to identify victims who have contacted the police. It is the combination of the granular level of detail, taken together with other ‘local knowledge’ or information in the public domain which can give rise to the risk. We also note, as a slightly different point that the very fact that information has (or has not) been reported can cause prejudice to law enforcement interests. Offenders can use the information in combination with other information that is available to them to avoid detection (i.e. if they know the offence has been reported due to the fact that a very similar/the same offence is recorded in the same spot they are likely to believe their offence has been reported, they may then choose to target victims in a different area, fearing the police may have patrols in the area or if the victim is known to them, they may threaten them), or to exploit law enforcement vulnerabilities. Noting that knowing the simple fact that an offence has not been reported can also lead to them being emboldened and perhaps committing a further similar offence as they are then aware that the police are unlikely to be targeting that area due to their lack of knowledge. We further noted that such identification (or fear of identification) could discourage victims from coming forward (noting that the underreporting of Serious Sexual Offences is well established).

139. We took into account Mr Greenwood’s statement and examples at paragraph 15 of his statement, as invited to by the Second Respondent. However, we noted his statement that “ Low numbers should only be redacted where this information would make it reasonably likely that the victim could be identified, and where the same information has not already been disclosed elsewhere (in the Met’s Crime Maps, in the media etc, see my first statement)”. The redactions that we concluded should remain redacted (expanded upon below) do fall into exactly this category.

140. We address the issue of low numbers and small data sets within the other headings. In light of the fact that we are unable to disclose the low numbers on some occasions (as this is the very information that is in dispute), we expanded upon the reasons in the closed schedule. (iv) Gaps and Recommendations

141. The Appellant submits its skeleton argument that: ‘44.4. Gaps and recommendations relating to specific BCU. This category relates to local policing methodology. While disclosure would provide some re-assurance on action taken to address the issues, this is outweighed by the need to protect police strategy and tactics. Disclosure risks informing offenders of policing methods used to combat crime and risks enabling them to exploit weaknesses and take evasive action. MPS has disclosed the vast majority of the gaps and recommendations information but seeks to withhold that which is too granular or relates to specific venues (eg. redactions at [C53, 55-57].’

142. The Second Appellant argues in its original skeleton argument that: “71…no real or actual prejudice can arise. Supt. Davies states in his evidence that this concern relates to “those areas where the Profile provides insights into methodology, tactics and activity that have been used, or are suggested for use, in combatting crime” [ B/233/§7(iv) ]. However, the approach the MPS have taken to this is entirely inconsistent.

72. The recommendations/intelligence gaps summary in the Profile is largely un-redacted [ B/257 ], and in nine of the BCU summary pages (AW, AS, SE, WA, SN, CN, CE, NE, EA) [ B/259-265 ], the ‘gaps and recommendations’ sections have also been largely disclosed. It is only on the NA summary page that the ‘gaps and recommendations’ section, has been redacted in full [ B/266 ]. This is inconsistent. The MPS cannot accept that no prejudice is suffered in relation to the disclosure of gaps and recommendation information in nine of its BCU, yet 1773. 74. somehow the same or equivalent information in one BCU purportedly does give rise to some kind of prejudice on the ground that that information somehow prejudices ‘police tactics’.

73. Similarly, in four of the BCU summary pages (AW, AS, SE, and SN) [ B/259-261, B/263 ], the ’long term ongoing issues and emerging trends’ sections have largely been disclosed, yet on the WA summary parge that section has been redacted almost entirely. Again, this is inconsistent. The MPS cannot accept that no prejudice is suffered in relation to the disclosure of issues and trends in at least four of its BCU, yet somehow the same or similar information in the WA BCU purportedly does give rise to some kind of prejudice to ‘police tactics’ such that that entire section has been redacted. In the context of the wider public debate around policing of sexual offences and violence against women and girls in London, these parts of the Profile which relate to long-term issues, emerging trends, gaps, and recommendations, are of crucial importance to the public interest, and no real or actual prejudice arises from their disclosure”.

143. We considered the ‘inconsistencies’ in detail. We do not address here why we find some recommendations/intelligence gaps have been disclosed and why others have not (because it is difficult to do so without referring to the closed information that has not been disclosed). However, we say for clarity, and in order to address one of the examples raised by the Second Respondent, that although it is correct that “Long Term On-going issues and Emerging Trends” sections have largely been disclosed, but on the WA summary page, that section has been redacted almost entirely. We do not agree this to be “inconsistent” we find that the information contained under the WA summary page redactions (internal page 24) contains highly sensitive information which is very different in substance to the information contained under the same heading in other sections.

144. We direct that the following redactions should remain redacted :

145. In terms of the first redaction on internal page 24 (spanning 16 lines, located in the first column), we concluded that this information was exempt under several sections. The Appellant relied upon s.31(1)(a) and/or (b) on the basis that “ Particular venues….and apps and relatively high in resolution maps enabling broad pinpointing of locations. Includes highly granular accounts of particular and repeat offences, reports and patterns (eg 24)….” (in relation to internal pages 21-24) . The redacted information contained very specific dates, highly detailed and particular descriptions of the circumstances, as well as the details of suspects and the language used. We agreed that if it was disclosed, it could lead to offenders becoming aware that they have been reported, their methodology had been disclosed and that they were on the police radar. This could result in a variety of consequences, including adaptions of methods and evasion of detection. It was submitted that the risk was of straightforward mosaic identification due to the high specificity. It would be easy to identify persons with part-knowledge e.g. a friend/colleague/associate who knows a little but not the full picture would get everything else they needed. We agreed that the redactions provide microscopic detail that if disclosed would result in a real possibility of prejudice to law enforcement objectives under s.31(a) and (b). We were under no doubt that there was a causal link between disclosure and the prejudice and that the nature of the prejudice that may follow disclosure was real and significant. Again, we considered the public interest test but concluded that maintaining the exemption outweighs the public interest in disclosure – in particular because even one offender gaining a material advantage would outweigh the public interest. We also agreed that the information was exempt under s.40 as the individuals would be identifiable from the redacted information, noting that even CRIS numbers are detailed here. We also found s.30 was applicable (in the alternative) for the reasons already specified.

146. The first two bullet points on internal page 26 (column 1, spanning 3 lines). It was not clear from what was before us, which exemptions the Appellant relied upon in relation to these redactions. Nonetheless, we found that s.31(1)(a) and (b) did apply because the redactions relate to tactics and intelligence. We concluded that disclosure could lead to offenders changing their tactics in order to avoid detection. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime, and (ii) the apprehension and prosecution of offenders. We were under no doubt that there was a causal link between disclosure and the prejudice and that the nature of the prejudice that may follow disclosure was real and significant, Again, we considered the public interest test and in particular that sometimes revealing strategies (for example revealing that the police conduct undercover drive-throughs (not applicable to this example but a submission made by the Appellant generally) would favour disclosure as it would reassure the public that the police are there even when they can’t see them) but concluded that the public interest in maintaining the exemption outweighs the public interest in disclosure in this specific case – in particular because even one offender gaining a material advantage would outweigh the public interest.

147. The sixth redaction on page 26 (column 2, fourth line): we concluded that this was quite specific information, which shows targeting of a particularly narrowly defined group in a particular area. We accepted that it was granular. This information taken together with other information on the same page provides a “strategy of offending” to offenders. Consequently, we found that both of the applicable interests in s.31 are relevant. We concluded there was a causal link between disclosure and prejudice for the reasons already provided and concluded that the likelihood of prejudice occurring is real and significant. Again we rely on the same public interests as previously. We found the balance tipped in favour of redaction, in particular because again even one offender gaining a material advantage would outweigh the public interest. We also found s.30 was engaged in the alternative.

148. The last redaction on internal page 26 – excluding the first three words (column 2, spanning 3 lines): we concluded that this should remain redacted under s.31(1)(a) and/or (b) as it contained information relating to police gaps which could encourage offenders to vary their tactics in order to take advantage of this information. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime, and (ii) the apprehension and prosecution of offenders. We were under no doubt that there was a causal link between disclosure and the real and significant prejudice that may follow. Again, we considered the public interest test but concluded that maintaining the exemption outweighs the public interest in disclosure – in particular because even one offender gaining a material advantage would outweigh the public interest.

149. The redaction on internal page 27: the Appellant relied upon s.31(1)(a) and/or (b). We concluded that it should remain redacted. We found that the first applicable interests in s.31 was relevant namely (i) preventing and detecting crime. We concluded that there was a causal link between disclosure and the real prejudice that may follow, as it involves disclosure of police intelligence. Again, we considered the public interest test but concluded that maintaining the exemption outweighs the public interest in disclosure – in particular because even one offender gaining a material advantage would outweigh the public interest.

150. The redactions in the second column on internal page 28: the Appellant relies on s.31(1)(a) and/or (b). The grounds for non-disclosure are “ Granular details of policing gaps and methodologies in relation to specific areas, groups or profiles” (in relation to the redactions at internal pages 25, 27-29). We concluded that these should remain redacted. We agree that the information contained therein is granular. It is very detailed and potentially gives offenders intelligence about where offences are reported and what knowledge the police have. It includes venues and addresses which were very specific with low numbers. The information set out under ‘Gaps & Recommendations’ refers to Police activity, techniques and tactics – including various forms of particular types of surveillance. We found the information to be highly granular and sensitive; we concluded that it could assist those who seek to avoid detection and prosecution. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime, and (ii) the apprehension and prosecution of offenders. We concluded that there was a clear causal link between disclosure and that real and significant prejudice may follow. Again, we considered the public interest test but concluded that maintaining the exemption outweighs the public interest in disclosure for the reasons already set out – in particular because even one offender gaining a material advantage would outweigh the public interest.

151. We concluded that all the redactions on internal page 29 should remain redacted. The Appellant relies on s.31(1)(a) and/or (b). The grounds for non-disclosure are “ Granular details of policing gaps and methodologies in relation to specific areas, groups or profiles” (in relation to the redactions at internal pages 25, 27-29). We concluded that the redactions relate to police tactics and tell the reader both what the police know about the offending, what they are looking for and where they intend to target their efforts. We concluded it was likely to prejudice either the prevention or detection of crime or the apprehension of offenders. We found that both of the applicable interests in s.31 are relevant: (i) preventing and detecting crime, and (ii) the apprehension and prosecution of offenders. We concluded that there was a clear causal link between disclosure and the real and significant prejudice that may follow. Again, we considered the public interest test but concluded that maintaining the exemption outweighs the public interest in disclosure – in particular because even one offender gaining a material advantage would outweigh the public interest. Costs

152. The Second Respondent requested Wasted Costs in the original skeleton argument. This was expanded upon in the most recent skeleton argument.

153. Paragraph 10(3) of the General Regulatory Chamber Procedure Rules states: ‘(3) A person making an application for an order under this rule must— (a) send or deliver a written application to the Tribunal and to the person against whom it is proposed that the order be made; and (b) send or deliver a schedule of the costs or expenses claimed with the application’. (4) An application for an order under paragraph (1) may be made at any time during the proceedings but may not be made later than 14 days after the date on which the Tribunal sends— (a) a decision notice recording the decision which finally disposes of all issues in the proceedings… (5) The Tribunal may not make an order under paragraph (1) or (1A) against a person (“the paying person”) without first— (a) giving that person an opportunity to make representations; and (b) if the paying person is an individual, considering that person's financial means. (6) The amount of costs or expenses to be paid under an order under paragraph (1) may be ascertained by— (a) summary assessment by the Tribunal; (b) agreement of a specified sum by the paying person and the person entitled to receive the costs or expenses (“the receiving person”); or (c) assessment of the whole or a specified part of the costs or expenses, including the costs or expenses of the assessment, incurred by the receiving person, if not agreed.

154. We are willing to treat the Second Respondent’s skeleton argument(s) as an ‘application’ in accordance with paragraph 3(a) of the Procedure Rules.

155. However, the Tribunal has not received a schedule of costs. As such the Second Respondent is directed to serve a schedule of costs or expenses (alongside any further written submissions he wishes to make) within 14 days of the date of this decision, if he wishes to purse this application.

156. The Appellant and First Respondent are permitted to serve any written response they wish to rely on. This must be done within 28 days of the date of this decision. We note here that we have a record of the oral submissions that were made at the hearing. Signed Date: Judge Kiai 10 th November 2025