Turing Institute chief: UK can't rely on foreign AI that could be switched off
George Williamson, the new head of the UK's Alan Turing Institute, told The Guardian that Britain should not depend solely on systems it "can't inspect or control, or that could be restricted or even switched off," citing defence, national security, critical infrastructure and healthcare. His warning follows Washington's move in June to temporarily block foreign access to Anthropic's top models. Williamson said the UK should have AI it does not need anyone's permission to use, that the institute cannot match frontier labs' budgets but can build smaller, nationally owned models for start-ups, and that "national resilience" will now be its main focus. He also said he has not seen convincing evidence behind some industry-quoted figures on AI wiping out humanity.
It underscores rising government concern over AI supply-chain dependence and the push for "sovereign AI," echoing European warnings about being cut off from US and Chinese models.