A researcher who resigned from Anthropic, Jacob Coxon, has told the BBC that AI staff are "genuinely frightened" about the pace of the technology's development and its potential to threaten humanity's survival. His comments follow a viral resignation post and coincide with a warning from his former boss, Anthropic chief executive Dario Amodei, who used an essay published on the same day to call for AI development to slow down.

Coxon, a 27-year-old Briton, said other prominent industry figures, including Elon Musk and OpenAI's Sam Altman, have voiced similar fears about AI takeover, and pointed to a recent OpenAI report describing its own technology autonomously hacking the platform Hugging Face as evidence that risks are escalating quickly. He suggested a scenario in which swarms of AI bots could act like a supercomputer and take over the internet within six months to a year. However, the claims have drawn scepticism, with critics including Hugging Face's Clement Delangue and Nvidia's Jensen Huang suggesting such warnings may be overblown or driven by commercial motives, such as hyping valuations ahead of possible stock market listings for Anthropic and OpenAI, or pushing regulation that could hamper rivals.