Source · Select Committees · Public Accounts Committee
Recommendation 11
11
Acknowledged
Public trust is vital for AI adoption but remains low, requiring greater transparency.
Conclusion
67% of government bodies responding to the NAO’s survey said that support from the centre of government in fostering public trust in AI was very important.23 Written evidence we received from a wide range of stakeholders emphasised the vital importance of public trust. For example, the Ada Lovelace Institute told us expanding the use of AI in the public sector without maintaining public trust and confidence could increase the risk of a public backlash or people withdrawing consent for the data to be used.24 The Local Government Association also raised concerns at survey findings that show low levels of public trust in how government uses the personal data of citizens.25 Transparency and openness were emphasised in the written evidence we received as important ways in which government can build trust, through meaningful public engagement and transparent reporting on AI use and evaluation.26
Government Response Summary
The government acknowledges the importance of public trust and data privacy in AI adoption.
Government Response
Acknowledged
HM Government
Acknowledged
2. PAC conclusion: Public trust is being jeopardised by slow progress on embedding transparency and establishing robust standards for AI adoption in the public sector. 2a. PAC recommendation: DSIT should write to the Committee in six months to update us on: • Departmental compliance with the Algorithmic Transparency Recording Standard and further action it is taking to tackle gaps in transparency to strengthen public trust, including to address public concerns over data privacy and the sharing of sensitive data. 2.1 The government agrees with the Committee’s recommendation. Target implementation date: January 2026 2.2 DSIT is continuing to implement the mandatory rollout of the Algorithmic Transparency Recording Standard (ATRS) in central government and is aiming to publish records for all currently identified in-scope algorithmic tools (as of March 2025) in government departments (not including ALBs) by the end of 2025. A range of activities to evaluate the effectiveness of the published ATRS records at strengthening public trust are planned to take place from summer 2025 and will be included in the Government Digital and AI Roadmap publication. 2b. PAC recommendation: • How its strengthened spend controls for high–risk AI use cases will support safe and ethical AI roll–out. 2.3 The government agrees with the Committee’s recommendation. Target implementation date: October 2025 2.4 In addition to the changes already made to the risk and importance framework to ensure that products and services that leverage AI are identified, GDS has also updated internal guidance for DSIT officials assuring AI spend. The spend control assurance of all Digital spend above thresholds that’s flagged as “AI” now checks for compliance with Algorithmic Transparency Recording Standard (ATRS). Further work is underway to develop an approach to assuring delivery on the back of that spend during digital service assessment exercises. An update will be provided in the Autumn by DSIT on any further changes to spend assurance, along with the changes introduced into service assessments as well. These initiatives will be part of the Government Digital and AI roadmap, planned to be published in Autumn 2025.