The Digital Regulation Cooperation Forum, which brings together the UK's main digital regulators, has been gathering views on how the public sees the risks of generative and agentic AI. The window for responses closed on 3 July.
Why agentic AI is the focus
Agentic systems act across several steps with limited human input, which changes the risk picture. A tool that drafts text is easier to supervise than one that takes actions, calls other services and commits an organisation to an outcome. Regulators asking about public attitudes now are mapping the ground before they decide where to press.
What firms can take from it
The direction is a fair guide to where UK scrutiny is heading, even without a single AI law behind it. A firm deploying agentic tools should be able to show where a human stays in control, what an agent may do on its own, and how its actions are logged. Those are the questions a regulator is likely to reach for first.
Keep a human in the loop, on the record
HEX 165 helps you govern agentic systems against the duties that apply. Get in touch or read more about HEX 165.