On October 31st, 2024, Demos, the Ada Lovelace Institute and Connected by Data convened a workshop of leading AI governance experts across civil society, industry, and government to discuss the forthcoming UK AI Bill – or, more precisely, to discuss what is happening around it.

The newly released AI Opportunities Action Plan goes some distance towards outlining plans specifically oriented toward promoting AI industry and AI adoption in the UK. However, as things stand, there remains a risk that if the frontier AI Bill is presented without parallel commitments and concrete plans for dealing with the array of harms and challenges posed by present day AI, that it may struggle to move through parliament. In turn, the Bill could fail to achieve its ambitions for frontier AI regulation and progress on wider AI regulation and policy would stagnate.

We convened our workshop to help articulate the spectrum of reassurances DSIT could offer to allay fears about the narrowness of the AI Bill, oriented around the following prompt:

What would we want to see happening alongside the Bill to provide assurance that other aspects of AI safety and societal and industry impacts are also being taken seriously and acted upon?
What would we need to see happening to be ok with the scope of the AI bill because we are satisfied that our broader concerns are being addressed by other means?

This briefing paper provides a summary of workshop findings as a supportive resource for DSIT, in planning for the announcement of the frontier AI Bill and for helping to clarify its broader AI policy strategy for the UK. We also hope it will prove useful for discussion across civil society, academia, industry and elsewhere.