LONDON, May 29 (Reuters) – British banks still have not been able to gain access to Anthropic’s Mythos artificial intelligence model to check their systems against cyber threats, six weeks after it first drew concern, Bank of England Governor Andrew Bailey said on Friday.
Bailey, speaking to Bloomberg TV, said Anthropic was willing to share the models on a trial basis but there appeared to be a political hold-up.
“It hasn’t happened yet and I think this has been somewhat caught up in the process with the U.S. administration,” he said in an interview on the sidelines of a central banking conference in Reykjavik.
“Quite why the process is a bit different from one company to another, I’m afraid I can’t explain to you. Obviously, from our point of view, given our concern about the risks involved in this, it’s very important that there is access,” he said.
Anthropic has sparred with the U.S. administration over guardrails for how the military could use its AI tools.
Last month, Bailey said “Anthropic may have found a way to crack the whole cyber risk world open”.
Since then, some cybersecurity experts have told Reuters that fears of unfettered hacking with the model are overstated.
President Donald Trump last week postponed signing a broader executive order on AI that was expected to create a voluntary framework for AI developers to engage with the U.S. government before the public release of advanced AI models.
Bailey – who also heads the international standard-setting Financial Stability Board – said there needed to be a global approach to hacking risks.
“Spillovers from this sort of cyber risk are so big that we can’t just have a single sort of national approach,” he said.
“Anybody who thought, ‘Well, I’ve dealt with my banks, that’s okay’, I’m afraid that won’t work, because they’re all so heavily interconnected.”
(Reporting by David Milliken; Editing by Daniel Wallis)
Comments