Corporate report

BFEG meeting minutes: 1 April 2025

Updated 9 September 2025

Notes of the 30th meeting held on 1 April 2025 online via MS Teams

1             Welcome, introductions and declarations of interest

1.1         The Chair welcomed all to the meeting and noted apologies. A full list of attendees and apologies is provided as annex A.

1.2         Members were invited to share any new or arising declarations of interest. New declarations of interest are noted below.

• Professor Sarah Morris noted she was undertaking two projects, one within the MOD, and another being and ACE commission.

1.3         Members were reminded it is their individual duty to determine whether a matter would constitute a conflict of interest and that it is the responsibility of the individual to declare these.

Minutes and review of actions

1.4         Minutes from BFEG meeting held in January 2025 were approved. One member had sent some grammatical changes, which will be circulated for final sign-off.

1.5         An updated action log was provided to attendees ahead of the meeting which included a status update against each action and suggested certain items were closed. No objections were raised regarding closure of these actions.

2             Chair’s update, including review of BFEG submissions

2.1         The Chair provided an update on recent events and progress since the last quarterly.

• Secretariat have met with the Policy Sponsor team regarding the Commissioning brief for the coming year.

• BFEG’s name change submission has been made to Lord Hanson and Minister Johnson

• Regarding recruitment of BFEG Chair, the Secretariat are working with the Public Appointments Team to progress this. It is expected the advert will be out in the next few weeks.

• Work is progressing with the HO Ethics Adviser to develop the HO Ethics Board including what role BFEG may play in this.

• The Home Office Research Development and Innovation Strategy, which BFEG was involved in providing ethical advice on, was launched at the Security and Policing Conference 2025.

• The Engagement an Impact briefing paper (which will be discussed further in the closed session) is the result of a new strategy being implemented and highlights some positive engagement results.

• The Framework of delivery which is the agreement document between BFEG as an arm’s length body (ALB) sponsored by our Minister Lord Hanson and is due for review. A draft has been submitted to the Minister by the 31/3/25 deadline as agreed.

• The Minister has indicated a clear desire for continued engagement with BFEG, and the Chair is due to meet with him in June.

2.2         The Chair attended the Security and Policing conference in March alongside some members and Secretariat. The Chair congratulated one of the members for reaching the final stage of the Academic Innovation Award, which was showcased at the event. Secretariat accompanied Minister Hanson’s private office around the event. Secretariat were also able to engage with other Arm’s length bodies and teams within the Home Office such as the Accelerated Capability Environment (ACE), which positively increased BFEG’s profile.

2.3         The Chair noted that Malcolm Oswald is leaving the BFEG and thanked him for his hard work and commitment to the panel.

Submissions

2.4         The chair gave an update on the submissions that BFEG have received since the previous quarterly meeting:

• Continuous engagement with His Majesty’s Passport Office, including a meeting to discuss the results from their research

• Updating ACS and APS pathfinder submissions​

• An urgent submission from (FINDS) regarding DNA contaminants​

• Public attitudes to the use of live facial recognition survey meeting​

• ACE Open-Source Intelligence (OSINT) Trials 2nd phase meeting​

• Initiating new ACE submission on synthetic data

• A BFEG member met with the Director General Home Office Ethics Champion after the January Quarterly meeting.

3             Policy update

3.1         Policy gave an update relating to Facial recognition (FR).

3.2         The survey assessing ‘Public Attitudes to Policing’s use of Live Facial Recognition’ is complete and waiting for publication. The initial analysis from the findings were:

• The results showed that the majority of people were in support of FR, particularly if it is being used in relation to violent and serious crime.

• People find the technology behind the FR and the legal framework in which it sits in either ambiguous or complicated.

• People were more likely to have concerns about FR if they lived in London, were from an ethnic minority, or identified as LBGTQI+.

• People were concerned about the misuse of FR and security risks (such as hacking).

• It is expected there would be a broader debate about police use of personal data, AI, and trust in police over stop and search.

3.3         It was highlighted that Policing leaders may be uncomfortable in leading the debate on what safeguards and constraints there should be regarding FR.

3.4         It was raised that there is also intrinsic bias in FR algorithms. It is important that this is fully understood.

3.5         Policy also noted that Facial images are a ‘biometric’ and must therefore be treated as a certain way in legislation and statute.

3.6         Some BFEG members joined a roundtable on FR. Ministers have now been brought a set of conclusions relating to the research.

3.7         A member asked for detail on how law reform will work in this area. It is suggested that BFEG may like to engage in this.

3.8         Policy officials indicate that work relating to this will go the Biometrics and Facial Recognition Working Group.

Action 0104_2025#01 Secretariat to send around information on each Working Group to the BFEG members.

3.9         An attendee of the meeting asked when the survey findings are likely to be published.

3.10       Policy responded that an estimate would be June, but that it will be agreed by Ministers.

3.11       A BFEG member asked about the possible extent of a change in legislation. Will there be a whole infrastructure of requirements and guidance for all the separately functioning bodies involved in this? Could BFEG play a role in this?

3.12       Policy responded that it would be expected that Parliament would want a code of practise of facial recognition. There will be a balance between what is put in primary and secondary legislation. Guidance currently in the College APP may move into the code of practise. BFEG could provide useful input in this.

3.13       Another BFEG member asked whether the Home Office would give interim guidance as to what type/seriousness of crimes FR technology could be used for?

3.14       Policy highlighted that they are recruiting the next Biometrics and Surveillance Camera Commissioner. They may then work with the Forensic Regulator and other regulators on a draft voluntary code of practise. The use of ‘serious’ is an ambiguous term and is not well defined in relation to crime type.

3.15       Members provided links in the chat to the definitions of ‘serious’ crime that have been given.

• Von Hirsch, Andrew, and Nils Jareborg. “Gauging criminal harm: a living-standard analysis.” Oxford J. Legal Stud. 11 (1991): 1.

• https://www.cps.gov.uk/legal-guidance/serious-crime-prevention-orders

• Braga, A. A., & Kennedy, D. M. (2021). A framework for addressing violence and serious crime: Focused deterrence, legitimacy, and prevention. Cambridge University Press.

•  https://global.oup.com/academic/product/eu-data-privacy-law-and-serious-crime-9780198837169?cc=gb&lang=en&# 

3.16       Another BFEG member congratulated Policy on the work that had gone into this survey and research. They ask if the scope of any new legislations would take into account emerging technologies.

3.17       Policy advised that they would like the independent Expert Oversight Body to have a power that enables them to draft codes of practise to address emerging technologies that may give stakeholders a cause for concern. FR recognition is a very unique type of data that should have its own statute. This may then be applied to other biometric technology. There is therefore a question as to whether potential legislation should be specific to ‘FR’ or should be slightly broader to encompass other similar technology. It would be important however that the specificity is not entirely lost due to the complexity of FR and requirement for it to have very detailed guidance and legislation surrounding it.

3.18       As a follow up, one of the members highlighted Article 8 of the European Convention of Human Rights and its implications on Biometric systems.

3.19       Policy explained the complexity of current legislation on any potential use of FR, referencing both article 8 and the EU AI act. These are questions and considerations that the government will have to think about.

3.20       In reference to the AI act, a member mentioned that in other countries, the use of FR has resulted in ‘mission creep,’ i.e., the technology is now being used in other ways that it was not originally intended for.

3.21       Policy responded that there would be a concern if this were to happen in the UK.

3.22       Policy followed that BFEG’s continued involvement in this will be important. There may also be a FR conference before summer. This may involve consultations and discussions with the public surrounding FR; however, it was noted that the individuals who may oppose FR may also be least likely to responded to government consultations.

3.23       A BFEG member highlighted a UKRI projected which had participation from individuals who may be usually absent from these discussions and are particularly worried about surveillance. The link is given below:

• https://digitalgood.net/dg-research/public-voices-in-ai/

4             Home Office Biometrics (HOB) update

4.1         A HOB Official provided a recent update of their work, which was sent round to all attendees prior to the meeting.

4.2         It was highlighted that one of the new areas of work is discussion around technology development. Whilst this would go to the Ethics WG, it could potentially be relevant to the Biometrics and FR WG.

Action_0104_2025#2 Secretariat to contact HOB Officials to determine what WG the new work may fall in to.

4.3         A HOB Official gave an update relating to fingerprints. Strategic match is the main focus of attention. This involves developing the platform upon which new fingerprint algorithms will reside on. Stage 1 is almost complete. This is in early parallel run with colleagues in DEAF bureau. The main parallel run will be up and running in May. This will be the largest collection. The other stage of delivery will be rolled out of the next 19 months. This includes 360-degree rotation of palms

4.4         The initial report on how matcha is progressing will be provided to the ethics WG.

4.5         Other aspects areas of work relating to fingerprints are around improving the tools that the police and immigration bureaus have.

4.6         There is a spending bid to take this work forwards, in with the outcome will be announced in a few months.

4.7         The members were also given an update regarding the contactless fingerprint search, which is close to getting research findings. This will be brough to the ethics WG in April/May.

4.8         A member asked whether there had been much progress in the work being undertaken by HOB since the last quarterly, as much of the information presented on the slides was the same.

4.9         A member of the HOB team responded that the work is progressing and is still ongoing. They noted that the retrospective facial recognition portal is now 80% complete instead of 50%.

4.10       Another member asked for a progress update on the Facial Estimation work that relates to the BFEG and AESAC reference group.

4.11       HOB officials offered to follow this up. The also suggest that this work may get brought to the ethics WG.

Action_0104_2025#03 HOB officials to contact secretariat about the FAE work being undertaken to schedule engagement with the joint reference group and the ethics WG.

4.12 A HOB official briefly updated the members of the work relating to DNA. This was sent around to attendees beforehand.

Action_0104_2025#04 HOB official to provide dates for the next ethics WG to meet.

5             Forensic Information Database Service (FINDS) update

5.1         A member from the FINDS team provided an update on the YSTR project.

• The project is progressing and is getting close to the 10,000 samples that they require.

• The £10 incentive has started.

• Phase B of the project.

5.2         An update was also provided on the current expansion of the DNA database, which is progressing.

5.3         A member from BFEG asked whether they were retaining samples, and whether all samples were being processed through YSTR. She also noted that it could be possible that a match is found on the database relating to a Y chromosome but not to an autosomal chromosome.

5.4         A FINDs official responded that samples are not retained, and that any sample take would be automatically processed through YSTR. In reference to the point about Y matches, additional quality checks relating to this have been proposed to the FINDs strategy board. This has been proposed for YSTRs where there may be commonality of profiles or male lineage. This would go to the forensic provider for them to assess the autosomal results. It was emphasised that matches like this would not go straight to the police but would be tested further to ensure the evidence is correct first.

5.5         Another member asked what prompted this quality control.

5.6         A FINDs official responded that this was due to current knowledge of YSTR profiling in the Policing community. It is important that Policing have a good understanding of this research if they are to use it. This is why it is important now that evidence goes to the experts first.

5.7         In reference to the YSTR, a member asked whether there were concerns relating to the reliability of the method in correct matches.

5.8         A member of the FINDs team responded that it is to understand the frequency of adventitious matches. They have been engaging with experts to model this. The YSTR method is not itself new in the UK, and so the method is well known and so are the limitations.

5.9         An update from FINDs was then provided on the incorporation of third-party software to carry out the searching and matching. They are currently looking at software called DBLR.

5.10       They also noted that they are currently producing a verification/validation pack of information that will be reviewed by and independent panel (the panel is yet to be created). This will help provide assurance that the data is performing as it should.

5.11       Considering this independent panel has not yet been created, A BFEG member asked whether they would then be performing a retrospective review.

5.12       FINDs responded that they would not go live with anything until the panel had approved its use. It will therefore not be retrospective.

5.13       A FINDs official also highlighted that two papers that had recently been submitted and that they welcome BFEGs review or comments. One of the papers related to 1134 DNA profiles that should have been loaded to the DNA data base of the past 12 years but were not. A gold group has been established to resolve this.

5.14       A member asked what the outcome of the work being undertaken by the gold group will be. If wrongdoing has been found, how will accountability be taken?

5.15       FINDS replied that it will be complex and that understanding why a backlog of DNA profiles being uploaded has occurred may be due to several reasons. Multiple stakeholders will need to be engaged with to ensure this does not happen again. In terms of accountability, it will also be a complex situation, and it may not be clear as to who holds accountability. More processes need to be put in place in the future to mitigate this and to also determine a chain of responsibility.

5.16       A BFEG members followed up by asking how the backlog was first discovered.

5.17       The FINDs team responded that they are tracked by the forensic service provider. This situation happened to one specific provider. The backlog was identified during an audit and was then escalated.

5.18       A member asked who the gold group was made up of and whether any external boards, such as the Information Commissioners Office (ICO), were involved?

5.19       FINDs responded that there are representatives from the forensic service prover, from FINDS, and from a large number from different policing regions. There is no one from the ICO. The group is led by policing, and the only oversight is the forensic science regulator.

5.20       Another member queried whether the use of YSTR could result in a miscarriage of justice, whether some who has been convicted should not be, or a victim who has not got justice.

5.21       FINDs replied that this is a major topic in the gold group. So far, however, this has not been seen.

5.22       A BFEG member went on to ask whether there had been an estimate of cost of this investigation, and whether it has been requiring substantial resourcing? They highlighted that at some point; someone may ask for this information

5.23       A FINDs official responded that they have not calculated this yet and agreed that it had required lots of resources. It would be possible to investigate this and provide an estimate.

Action-0104_2025#04 FINDs officials to investigate the cost and resourcing that has been required to run the investigation into the backlog of DNA profile uploads into the database.

5.24       BFEG members asked what the procedure is if a miscarriage of justice occurs relating to the use of YSTR, especially considering the incident that has occurred regarding the backlog of data inputs. Would it end up in the public domain, and/or is any of this already public knowledge? And as a follow up, BFEG highlighted that there is a duty of candour and a level of openness regarding this incident that has occurred.

5.25       A FINDs official responded that the Chair is considering this topic, including what information may need to be released, and that they would follow up with this.

Action_0104_2025#05 FINDS team to follow up with the question from BFEG as to what the procedure is if there is miscarriage of justice relating to the use of YSTR.

6             NPCC Update

6.1         A short update on the use of AI in Policing was provided from an official from the NPCC.

6.2         This included NPCC priorities relating to AI use and how AI technology adoption can be accelerated. There are three strategic priorities:

• Enhancing productivity

• Using AI to tackle crimes and harm

• Countering the criminal threat from AI

Action_0104_2025#06 NPCC Official to provide PowerPoint slides to BFEG members after the meeting.

6.3         An official from NPCC explained that to embrace these new technologies, this would require centralisation of capabilities, particularly digital technology, and procurement. Whilst this may take 10 years, the requirements to embrace these technologies will occur sooner.

6.4         A member from BFEG asked what the effects of centralisation will be. It was highlighted that BFEG has a historical interest in trying to advise the police on AI projects. However, because there are a lot of stakeholders involved in this process and the timetable for delivery has slipped considerably, it has been challenging to progress any work in synchrony.

6.5         The NPCC official responded that they are aware of these frustrations, and that they are looking to work with the national data analytics office to design a co-funding to support the progression of this.

6.6         It was highlighted by a member of the NPCC that at the moment, if someone within the Policing community lacks understanding on the ethical, legal, or technical aspects relating to AI, then they have no one to go to about this. There needs to be framework in place to rectify this.

6.7         This was re-iterated by a member of BFEG that is supporting this work on a Policing committee in the West Midlands.

6.8         Another BFEG members asked, in reference to an LLM event shared on the PowerPoint, whether they covered challenges with large language model, hallucinations, and other issues.

6.9         It was confirmed that issues like this were covered in the event.

6.10       A member of BFEG noted that similar work to do with an AI checklist may be being undertaken by the AI WG from BFEG, and that they should check that no duplication of work was being done.

6.11       Due to time constraints, it was suggested that any further questions relating to AI in the NPCC are picked up at a future point.

Action_0104_2025#07 Secretariat to arrange a follow up meeting with the NPCC to continue the discussion at the quarterly meeting.

7             Operation Tabula

7.1         BFEG were informed that there had been significant progress regarding the retention regime for custody images.

7.2         The key update points given were:

• It appears that the ministerial ambitions are that there is a progression of legislation, this is waiting to be confirmed

• The team is making sure they are aligning to the government ambitions

• It is likely that they will be applying the protection of freedom act to custody images to align to other biometrics

• They will be asking the Chiefs council in the summer to implement a regime aligning to any legislative changes to custody image retention

• The team have gone out to several police forces to determine whether this change will impact them. The majority did not feel that it would, as the protection of freedoms act is already applied to DNA and fingerprints

7.3         A member of BFEG asked whether the regime would be covered by the Biometrics and surveillance camera commissioner review.

7.4         The policing official responded that they would look at what the legislation needs to be to align custody images with other biometric date. This will involve working with the Office of Biometric and Surveillance Camera Commissioner on section 63G. It was reinforced that they are still waiting for ministerial agreement. For the records that would fall outside of the freedom of information act’s rules, the section 63G process would apply. They will also look into designing a review process with Police to address historic records appropriately.

8             Update from the College of Policing

8.1         An introduction to the College of Policing (CoP) was provided to BFEG members.

8.2         The College of Policing sets standards for policing and provides professional development and shared knowledge.

8.3          An update was provided on the ‘Centre For Police Productivity.’ The centre was formed to help forces roll out rapid innovation and to help measure the time-saving effect of this innovation. A large part of the work is to spot innovations and test/replicate them. The other side of the work is to provide a diagnostic tool for different forces to compare each other’s productivity. They can then find ways to increase productivity if needed. To do this, they are trying to solve a large ‘data sharing’ issue that exists as there is no centralised data management solution.

8.4               The official from CoP highlighted that AI is a major innovative technology. CoP is aiming to provide a set of practise guidance relating to the building capability and evaluation of building AI. They also have data ethics authorised professional practise, and data driven technology APP. These documents use BFEG principles in their development.

8.5          A BFEG members highlighted that a lot of technology that is considered ‘AI’ is not actually ‘AI.’ They asked whether a definition of AI was provided when Policing teams were asked to consider their AI usage.

8.6          A CoP official responded that a definition is given. They also added that when examples of AI are brought to them, they provide internal evaluation, advice, and review.

8.7          Another member asked whether they could expand (in reference to the slide) on the procurement of emerging technologies, and explain the risks associated with this.

8.8          CoP responded that purchasing is only one of a whole set of documents. It has been written specifically for people involved in procurement. It sits underneath the data driven technologies authorised professional practise. There is a centralised place where someone can find all these documents, including the responsible AI checklist. To share technology, there are two different networks working on this. They bring people together for knowledge sharing.

8.9         The member followed up as to whether there was anyone with legal capacity informing these reviews.

8.10       It was confirmed that at the moment, there is no one with legal expertise at specialised level.

8.11       Another BFEG member was concerned as to whether ‘efficiency’ was the best metric to assess innovative AI technology.

8.12       CoP responded that ‘efficiency’ is not the entire focus. The productivity focus is trying to get the same level of service for a better cost, or a better level of service.

8.13       A BFEG member asked that within the resources and guidance’s that they have, whether they also have any detailed impact assessments? Although Police may not legally be required to do this, there may also be some real benefit in doing it as it asks very searching questions to the people filling in the documents. The templates that have open questions as opposed to checklists are best.

8.14       A CoP official responded that although they do have a lot of documents, they do not have something as detailed as BFEG suggested. They suggested that they would take this point away.

Action_0104_2025#08 Secretariat to schedule a meeting between CoP and BFEG AI WG to continue discussions. This could be a joint meeting with the NPCC officials who also attended the quarterly.

9             Forensic Pathology overview

9.1         An official from Forensic Pathology gave an overview of Home Office forensic pathology.

9.2         It was explained that in England and Wales, the provision of the medical identification and assessment if death is not state provided. There are six group practises that cover a geographic area. The Home office impose rules and codes of practise which determine standards of forensic autopsies.

9.3         The team explained the difference between a hospital pathologist and a forensic pathologist and highlighted some of the challenges and difficulties facing the forensic pathology profession. This includes the structure of the profession, as they each act differently and are only governed by the rules imposed on them by the HO but nothing more. Another issue is that they are not associated with the National Health Service (NHS), which means that doctors do not have to, and may not often be willing, to assist in police cases that require specialists (often involving babies).

9.4         They went on to explain that due to the way that funding works per case, there is a disincentive to recruit forensic pathologists to group practise as it reduces the income and workload per person.

9.5         Over recent years, the Home Office Forensic Pathology has taken responsibility for most issues related to guidance for police in relation to the medical aspects of the investigation of death and matters related to human tissues within the criminal justice centre. Previously, an ethics group had been sent up relating to the retention of human tissues.

9.6         A BFEG member highlighted the complexity of repatriating human remains and how traumatising it can be for the individuals who must go through this.

9.7         Officials from the HO Forensic Pathologists agreed that the most important aspect of this work is to put the individuals and the families first in these situations. They explained that they had changed the Police practise advice to ensure that the family are fully appraised (to the extent that they want to be) on what the authorities find.

9.8         A member of BFEG expressed their concerns in the fact that forensic pathology is not covered by the remit of the forensic regulator. Therefore, there are concerns about the quality of the processes that are undertaken in the guise of forensic medical pathology.

9.9         HO forensic pathology officials responded that there is a clear distinction between forensic science and forensic medicine, which explains why they separated from the forensic science regulator. Now they regulate forensic pathology themselves through the pathology delivery boards. Forensic pathology is subject to a code of practise, and in Scotland they use a variant of this code. The processes of overseeing forensic pathology have been similar since 1990.

9.10       A BFEG member questioned whether these then met the quality standards in relation to evidence and interpretation that are seen in forensic science, and whether it is as rigorous as it could be? They also expressed concern at the categorical opinion that are often stated by forensic pathologists, whereas new standards are being written for forensic science to avoid categorical opinions in all cases.

9.11       A HO forensic pathologist official highlighted that they follow the same procedure as many other medical professions. In terms of rigor, they ensure that through their processes, they are as robust as they can be. They welcomed any measures that could increase robustness. Because forensic pathologists have to give their opinion into the cause of death, this is often a statement. Therefore, it can be hard to give a non-categorical opinion in this matter.

9.12       Another member asked if there was a specific ask of BFEG that the team may have for them.

9.13       The HO forensic pathologist officials replied that they do receive ethical issues relating to human tissues that they would be happy to bring to the committee if the police required an opinion or resolution. Many tissues samples are very small samples on glass slides; therefore, it can be ethically challenging to know how they should be treated in regard to the Human Tissue Act. They did highlight that most of their work is operational.

Action_0204_2025#09 Secretariat to contact the forensic pathology team to discuss whether they may need any input from BFEG.

10          Closing remarks and AOB

10.1       No other business was declared.

10.2       The date for the next meeting (which will be the in-person meeting) to be agreed during the closed session.

Annex A – Attendees

Attendees:

BFEG:

Mark Watson-Gandy (Chair)

Richard Guest (attended until 16:00)

Giles Herdale

Matt James

Dave Lewis

Sarah Morris

Nora Ni Loideain

Niamh Nic Daeid

Marion Oswald

Malcolm Oswald

Charles Raab

Tom Sorell

Denise Syndercombe-Court

Peter Waggett

Elisabeth Mackay

Officials:

BFEG Secretary and Secretariat

Home Office Deputy Chief Scientific Advisor (attended until 15:00)

Data and Identity Policy Representatives

Representatives from the Home Office Biometrics Programme

Representatives from Forensic Information Database Service

Representative from National Police Chiefs Council

Representative from Police Digital Service

Representative from College of Policing

Representatives from Home Office Forensic Pathology

Apologies:

BFEG: Anne-Maree Farrell, Penney Lewis