London, United Kingdom. Britain must develop its own AI infrastructure, models and skills to avoid full reliance on US technology, according to Harriet Rees, a government-appointed adviser to the banking sector.
Rees said limited access for UK banks to Anthropic’s AI model Mythos underscored the urgency of a more strategic approach to artificial intelligence.
Call for AI capability
Rees, Starling Bank’s chief information officer, was appointed as the finance ministry’s “AI champion” in January.
“Now is the time for us to think very strategically about what we need to do to protect our leading position moving forward. Time really is of the essence; we don’t have two years here,” Rees said in an interview with Reuters.
Limited access to Mythos
US-based Anthropic released Mythos in April to a select group that included US lender JPMorgan.
Banks are seeking access to Mythos because it is considered an advanced AI model for identifying cybersecurity vulnerabilities, allowing them to address weaknesses more quickly and strengthen defences.
Only a small number of banks in Britain, mainly UK operations of US lenders, have been granted access, Rees said. The chief executive of one of Britain’s largest banks said last week that major British lenders did not have a timeline for access to the model.
US restrictions and wider rollout
The US administration introduced strict restrictions on access to Mythos. At the end of June, Anthropic said it had received permission to release the model to trusted US organisations.
An Anthropic spokesperson said the company had begun rolling out Mythos 5 to organisations outside the United States and was continuing to coordinate with the US government to expand access to domestic and international partners.
Bank of England Governor Andrew Bailey renewed his call for the United States to take a more cooperative approach to addressing risks from frontier AI models in his Mansion House speech on Tuesday evening.