EU Says AI Firms Must Prove Services Safe Before Operating in Bloc

The European Union supports the development of artificial intelligence but expects companies to demonstrate that their services are safe for citizens before they can operate in the bloc, a European Commission spokesperson said on Monday.

Thomas Regnier, speaking for the Commission, set out the position as the debate over the pace and risks of AI returned to the spotlight.

The discussion was reignited after Dario Amodei, chief executive of the AI company Anthropic, said on Saturday that firms must slow the pace at which AI models are advancing.

Amodei's remarks added fresh momentum to an ongoing conversation among policymakers and industry leaders about how quickly increasingly capable AI systems should be developed and released.

The Commission's stance places the burden of proof on companies: innovation is encouraged, but safety assurances for citizens are a precondition for operating in the EU market.

The comments come amid wider international scrutiny of AI development, with regulators weighing how to balance technological progress against public protection.