Article by Sarah O’Connor and John Burn-Murdoch: “In the UK, the beleaguered Labour government has high hopes that AI can help to deliver better, faster and cheaper public services. Just this week, the FT’s public policy editor Chris Smyth reported on a draft proposal to scale back recruitment plans for the NHS, which is partly premised on the expectation that much wider use of AI can boost productivity and even “completely substitute” for some (unspecified) roles.
But how easy is it to insert AI into public services? And how will we know if it’s working?…
Here in the UK, adoption of AI tools is gathering pace across many parts of the public sector. As far back as early 2024, the National Audit Office found that almost three-quarters of government departments and national bodies were either deploying or piloting AI tools. Some of these deployments are very narrow. The Department for Transport, for example, is trialling an AI tool to prevent fraud when people apply for subsidies for electric vehicle chargers, by checking “proof of installation” photos against similar or stock photos. Others are more widespread and seem to have developed quite organically, such as the rapid adoption of AI transcription tools by social workers to record meetings and save time typing up notes.
But how do we know if any of this is making a difference when it comes to productivity?..(More)”.
