AI development should be held back while governments and businesses address the risk of frontier models – such as Anthropic’s Mythos 5 – that can act like super-powered hackers, the UK’s former spy chief has warned. “It’s not just Anthropic. OpenAI is in this space, and Chinese frontier models will be in exactly the same space. We have got to find a way of slowing things down just enough to be able to work through the safety angles of this,” said Richard Moore, former head of the UK’s Secret Intelligence Service – better known as MI6. Moore now acts as…
News Timeline:
Track the development of this news story across the Internet.