Turing Institute head says UK should not be dependent on foreign AI

George Williamson, head of the UK's Alan Turing Institute, said the UK should not rely on foreign AI systems that could be shut down at short notice, and needs to develop its own version of the technology, making national resilience a priority for the institute. He said AI is increasingly embedded in defense, national security, critical infrastructure, and healthcare, and the UK should not rely solely on systems that cannot be inspected, controlled, or could be restricted or even shut down. Williamson mentioned that in June this year, Washington temporarily blocked the release of Anthropic's latest model outside the US, and there have been reports that the US has asked major AI labs not to give models to the UK AI Security Institute for testing before US counterparts have inspected them. He said the Turing Institute will focus on protecting critical systems from disruption, addressing threats from AI itself, and helping the UK develop appropriate strategic AI capabilities, adding that the institute could produce nationally owned intellectual property or smaller models for startups to use.