Transparency data

Centre for Data Ethics and Innovation Minutes - Board Meeting 1

Updated 12 October 2023

This transparency data was withdrawn on

This content is no longer current. The Centre for Data Ethics and Innovation (CDEI) Advisory Board closed on 9 September 2023.

Monday 10 December

10am - 2pm Tech UK

10 St Bride St, London EC4A 4AD

Attendees:

  • Roger Taylor (RT)
  • Rt Revd Dr Steven Croft (SC)
  • Edwina Dunn (ED)
  • Prof Luciano Floridi (LF)
  • Dame Patricia Hodgson (PH)
  • Richard Sargeant (RS)
  • Kriti Sharma (KS) (by phone)
  • Dame Glenys Stacey (GS) (Board appointment starts March 2019)
  • Dr Adrian Weller (AW)
  • Prof Lord Robert Winston (RW)

Apologies:

  • Dr Susan Liautaud
  • Baroness Kate Rock

Guest speakers:

  • Sir Nigel Shadbolt (NS)
  • Prof Nello Cristianini (NC)

Observers (CDEI staff and consultants):

  • Jen Boon (JB)
  • Ollie Buckley (OB)
  • Sam Cannicott (SC)
  • Isabella Cheung (IC)
  • Natalie Davis (ND)
  • Jonathan Dimson (JD)
  • Stephen Dunne (SD)
  • Alex Green (AG)
  • Alex Lawrence-Archer (ALA)
  • Lara McDonald (LM)
  • Foluke Oshin (FO)
  • Louise Pakseresht (LP)
  • Tom Youldon (TY)

Item 1: Welcome and Overview of Agenda

The Chair, RT, welcomed and thanked the Board, introducing Antony Walker the Tech UK Deputy CEO. Antony Walker expressed Tech UK’s support of the CDEI, highlighting Tech UK’s interest in the CDEI’s work areas, the annual Tech UK Digital Ethics summit and offering future hosting. RT gave an overview of the agenda. The Board confirmed there were no declarations of interest in relation to the agenda. The Board agreed to the audio recording of the meeting to aid in the creation of meeting minutes, noting that the recording would be securely destroyed following this process.

Item 2: Board Ways of Working

At the Board induction day, there was a discussion about how the Board would work together and with the CDEI team. A paper on ‘Ways of Working’ was circulated ahead of the meeting.

ND gave an overview of the paper, noting the following points:

Collective responsibility and public commentary:

  • Board meetings should be an open environment that allow discussion and disagreement. Final Board decisions should be supported by all Board members. Board members who, in another capacity, wish to express views publicly that are relevant to CDEI issues but not aligned to CDEI views should notify the Chair/Secretariat in advance

Number of meetings a year:

  • There will be eight Board meetings in 2019, which may comprise meetings, strategy days and public engagement sessions. Dates for these have been confirmed with Board members.
  • Board papers will be distributed with a minimum of one week for reading prior to meetings.

ND gave an overview of the paper’s recommendations:

Decision-making:

  • ND outlined the proposals around decision making, highlighting the recommendation that a quorum of the Board would be two thirds and decisions would be approved by a simple majority of the quorum.
  • ND proposed a workshop session at a future Board meeting to determine what levels of decision making will be brought to the Board.

Decisions:

  • The Board agreed with the proposed decision making process.
  • The Board agreed a session to work through levels of decision making at a future meeting.

Transparency:

  • ND highlighted the importance of transparency, noting the government’s consultation response which stated: “board meetings to be open to the public as far as possible”. ND reflected on a previous Board discussion in which the Board was clear they wanted the CDEI to be transparent in a meaningful way.
  • Options raised in discussion included: a mix of open and closed Board papers; live broadcast of selected meetings; publishing papers and minutes; opening selected meetings; and opening segments of meetings.
  • The need for the Board to have a sufficient comms plan prior to public engagement was highlighted.

Decisions:

  • The Board agreed that Board papers would be closed to ensure the Board could receive candid advice and sensitive information if necessary. Agendas and minutes would be published.
  • The Board agreed that around two meetings a year would have a structured open session.

Location:

  • The national remit of the CDEI was also raised, with suggestions to hold meetings in locations around the UK.

Decisions:

  • The Board agreed to hold some meetings outside London, highlighting Edinburgh, Manchester, Northern Ireland and Wales as potential options.

Role of Board members outside Board meetings:

  • Opportunities for members to get involved in developing CDEI work outside of Board meetings were discussed. These included: involvement with projects on targeting and bias; developing the Analyse and Anticipate function; developing a principles framework; the public engagement and communications strategies; and the future work of the CDEI.

Actions:

  • RT would agree with individual Board members the areas of work they would each be involved with.

Item 3: Strategy

The CDEI is developing a strategy with a view to publishing it in spring 2019.

Overview of Strategy:

The Board were asked for comments on the skeleton strategy document, noting that a full draft of the strategy will be shared prior to the next Board meeting. The Board highlighted the following in discussion:

  • Need to address outcomes we hope to see and help bring about through the CDEI’s work in addition to measuring direct outputs, acknowledging that these may only be partly or indirectly attributable.
  • An emphasis on the importance of public engagement and learning lessons from public acceptance of GM foods, nuclear waste and the MMR vaccine.
  • A positive reception for the ambition of global, rather than just national, influence and of tracking developments globally.
  • The importance of not duplicating work being undertaken by other bodies.
  • The importance of working with regulators.
  • The need for a section addressing accountability to government and planning our trajectory towards an independent statutory body.

Actions:

  • Board members to feed through any further thoughts to the CDEI team.

Strategic Focus and Operational Objectives:

  • The importance of language was highlighted, using positive framing of benefits and accessible terminology as far as possible.
  • Policy work should be engaging and understandable for the public and designed to enhance public debate. It should not be focused solely on policy audiences.
  • Discussion of algorithmic decision-making needs to include consideration of its data context, noting that data and algorithms cannot be examined separately when considering their impacts.
  • The CDEI focus on promoting innovation should be targeted on areas where ethical and / or governance concerns are a barrier to innovation and where changes to governance could open up responsible innovation opportunities. It should also look at areas where innovation in AI can help address existing ethical problems.
  • The CDEI’s five draft operational objectives were discussed:
        1. Establishing mechanisms for monitoring opportunities and risks;
        2. Demonstrating practical interventions in the two focus areas of targeting and algorithmic bias;
        3. Building awareness of the CDEI and developing its brand;
        4. Implementing a sustainable model for engagement for public, policymakers and international stakeholders;
        Putting operational support structures in place to deliver the CDEI’s work.
  • The value of developing clear success measures for these objectives was discussed along with the difficulty of both generating and measuring impact on the wider public. It was suggested that the CDEI develop numerical targets to demonstrate its intended impact.
  • The need for the CDEI to retain reactive capacity was discussed along with the value of developing a brand.

Decisions:

  • The Board broadly agreed with the strategic focus, subject to the comments noted above.

Actions:

  • CDEI staff to further develop language used in strategy.
  • Draft of strategy to be discussed at the next Board meeting.

Incorporating a framework of principles into CDEI's Work:

  • OB reflected on a previous Board discussion on the value of developing an underlying framework of principles for the CDEI’s work.
  • The use of principles as primarily an evaluative tool was discussed.
  • The opportunity to build on prior work was highlighted, for example distilling existing frameworks and reviewing work in bioethics and broadcast regulation.
  • Principles should be framed in a simple and accessible language, with clarity as far as possible about the practical implications.
  • It was suggested that alongside any framework of principles there should be an additional commitment to ensure that the application of principles should be assessed against the normative values of society.
  • The principles should not be prioritised in the abstract. Any weighting and trade-off between them should be considered in context-specific setting.

Actions:

  • The Board agreed that the CDEI Strategy should consider including a framework of principles which builds on existing thinking.
  • The Board agreed that a Working Group of Board members would be formed to further consider ethical frameworks.

Prioritisation Framework:

  • The Board discussed the framework proposed by CDEI staff to support the Board to decide which opportunities and risks to explore further.
  • They noted that the framework proposed could provide a useful basis to identify a range of high-potential opportunities and risks to explore further. The issues identified were sensible although the wording used to describe some needed to be reviewed.
  • They discussed the challenges of determining the relative importance of different opportunities and risks - including how to establish which factors are more important and how to measure those which are not easily quantified.
  • There was a debate about the value of attempting to quantify specific issues as opposed to considering the different issues in the round and forming an overall narrative judgement. If quantification was used, the process used would need to be sufficiently rigorous and clear.
  • A more explicit inclusion of the degree of public concern around each opportunity and risk was suggested as an additional factor.

Actions:

  • The Board agreed that a prioritisation framework is likely to be a useful tool for initial assessments of which opportunities and risks the CDEI should prioritise, as well as for documenting and communicating how the CDEI team has determined priorities.
  • CDEI staff will further refine the framework and test it through its application to the targeting project area.

General Approach to Projects:

  • OB outlined the proposed general approach to projects, including an emphasis on collaborative working, publishing outputs throughout project delivery and drawing from a broad range of expertise by supplementing internal resource with external expertise.
  • The Board discussed the importance of adopting suitable methodology and governance which might vary across projects, the risk of delays to project delivery resulting from the complexity of the issues being addressed, the proposed collaborative approach and the need to retain reactive capacity in the CDEI.
  • The importance of building time into projects to ensure that desired outcomes are achieved was highlighted, for example supporting and monitoring the uptake of recommendations following publication.

Item 4: Projects (Targeting and Bias)

JB presented the two project areas the CDEI has agreed with government, highlighting existing activity by civil society and policy organisations along with potential partners and possible models for engaging with them.

Targeting - next steps

  • Will follow a relatively traditional policy development process, building on academic work.
  • Will include a significant element of public deliberation with the opportunity to explore nuance around opinion.
  • Will seek to develop potential solutions, which might include traditional policy or regulation recommendations, but will also look ahead at the research and development pipeline of technological solutions to consider what might be needed to enable future solutions.

Bias - next steps

  • Will follow a modular approach with multiple areas of focus. Financial services, local government and recruitment have already been identified as potential areas of focus.
  • Further areas of focus might include justice or a more culturally-oriented area such as streaming media services or online dating.

Project Initiation Documents will be brought to the next Board meeting for approval.

Item 5: Budget for Remainder of 2018/19

  • OB outlined the CDEI’s remaining budget for 2018/19, identifying committed costs.
  • The Board noted the one-off spend on consultancy to support the establishment of the CDEI, during the Nov-Dec 2018 period.
  • The Board was invited to provide input on a programme of proposed activities including:
    • commissioning public engagement activity to provide input to project work;
    • engaging a communications agency to manage the design and build of an independent website for the CDEI (dependent on approval from the Government Digital Service), and to support the development of the CDEI brand; and
    • research support from the Cabinet Office Open Innovation Team and commissioning research externally on an additional area of interest, for example the data brokering industry.
  • The Board discussed the proposed activities, highlighting the importance of carefully defining research questions before commissioning work, being aware of existing work undertaken by others, particularly when considering public engagement.
  • The Board agreed that the CDEI should have an independent website, highlighting the importance of its design as well as the broader value of storytelling in translating research and engaging the public.

Actions:

  • The Board agreed the proposed spending plans.
  • The CDEI team will further develop work on a public engagement strategy.
  • Plans for a public engagement exercise, potentially delivered through Sciencewise, will be reviewed by the steering group of the targeting project.

Item 6: Sir Nigel Shadbolt - Data Sharing

NS gave a presentation to the Board, describing the mission, vision and history of the Open Data Institute (ODI). This provided an overview on the potential value of open data and ethical data sharing frameworks as a public good; and specifically addressed the ODI’s current work on defining and piloting data trusts. The three pilots ODI are considering are in: Local Authority urban data; plastics; and illegal wildlife trade. NS also discussed ethical priorities including the monitoring of children’s apps and advertising in insurance, lifestyle analysis and physical activity. The Board discussed the cost of establishing ethical mechanisms to enable access to public data including the public cost of maintaining the necessary infrastructure. The Board agreed that infrastructure provided a useful metaphor for key data - this is particularly helpful as it raises issues such as standards; current areas of concern in tracking and privacy; and how to integrate consideration of ethics into data collection systems.

Item 7: Prof. Nello Cristianini - Technical Expertise

NC is Professor of Artificial Intelligence at Bristol University. He briefly discussed his key concerns regarding current technological developments and put forward suggested questions the CDEI should be asking as it scopes its projects on targeting and bias.

Item 8: Closing and Next Steps

  • The Chair recapped that Board members would be able to join working groups to contribute to the development of specific areas of the CDEI’s work.
  • The next Board meeting is scheduled for 31st January 2019.
  • AOB - no items were raised.