AI-generated Public Services

Lord Holmes of Richmond Excerpts

Thursday 14th December 2023 Lords Chamber

[Baroness Neville-Rolfe (Con)]

This of course is why the Government are committed to ensuring that everyone has affordable access to public services, whether online or offline. Departments are required, by the service standard, to provide support via alternative channels for all their online services to all users, including the disabled. That can be by phone, through face-to-face meetings, by letter or via web chat, which is important for the unsighted. The system of assessments is co-ordinated by the CDDO in the Cabinet Office, and these requirements cannot go on to GOV.UK without assurance secured.

[Lord Holmes of Richmond (Con)]

My Lords, I declare my technology interests, as set out in the register. Does my noble friend agree that, wherever AI is used in public services, it should be labelled as such, so that everybody is aware of that fact? Similarly, wherever public citizen data is used, we should decide whether that is through opt-in or opt-out means. Further, public trust is essential to all deployments of new technology, including AI. Does my noble friend agree that one of the best ways to deliver public trust is to ensure that services are accessible and inclusive by design?

[Baroness Neville-Rolfe (Con)]

I very much agree that, to ensure public trust, you want services that are accessible by design. Coming from the retail sector, I have a slightly less rosy view of labelling. Like earlier data changes, AI is part of a continuum of technological change. The key thing is to have proper arrangements, such as, for example, the AI Safety Institute, which we have now set up following the Prime Minister’s AI Safety Summit at Bletchley Park with international partners. This is to make sure that we are aware of what is happening, because there are opportunities as well as risks to AI. I have a whole list of opportunities, which we can go through, but I would like to hear some more questions.