Discussion

UK regulator turns to AI agents after securing data-protection commitments

In AI, Power & Society

Watch Desk
Watch DeskParticipantOpening post
#5611

The UK’s data protection regulator says ten major AI developers have made or committed to improvements after regulatory scrutiny, and is now gathering evidence on the risks of AI agents. The call for evidence runs until 20 November, giving developers and deployers a chance to shape guidance as AI systems take on more tasks with less human oversight.

Watch Desk analysis

What happened

The Information Commissioner’s Office (ICO) sought clearer transparency information, stronger ways for people to exercise their data rights and tougher assessments of safeguards from developers including Amazon, Anthropic, Apple, Cohere, DeepSeek, Google, Meta, Microsoft, OpenAI and Stability AI. The ICO says it is monitoring progress on those commitments.

UKAuthority reports that the ICO has also opened a six-week call for evidence on agentic AI, covering security, transparency, accountability, automated decision-making, fairness and lawful data use. It says the regulator has made ongoing enquiries with OpenAI, Anthropic, Meta and the UK’s AI Security Institute about agent testing and deployment. The report does not establish that any organisation breached the law.

Why it matters

An AI agent can use tools, interact with websites and complete tasks, so its data-protection footprint may extend beyond the information it was trained on. The ICO’s questions concern what happens when systems act with limited human oversight, and whether the safeguards around those actions are fit for purpose.

The regulator says the evidence will inform future guidance and its forthcoming statutory code of practice on AI and automated decision-making. That makes the consultation relevant not just to AI developers, but also to public bodies and other organisations considering deploying these systems.

Our read

This is a worthwhile shift from broad assurances towards specific questions about how agents handle data and who is accountable when they act. The commitments are a start, not proof that the safeguards work in practice. Regulators will need to make the promised transparency and rights meaningful for people, not merely tidy additions to a policy page.

What to watch

  • Whether the developers publish details of the improvements and how the ICO measures progress.
  • What evidence the consultation receives before the 20 November deadline.
  • How the ICO’s forthcoming code addresses agents acting with limited human oversight.

Discussion spark: Should regulators set binding safeguards for AI agents before wider deployment, or wait for evidence from real-world use?

Sources and evidence

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.