Ex-Anthropic researcher's viral AI warning breaks through

A 27-year-old former Anthropic pretraining researcher named Jacob Coxon resigned on September 8 and posted a warning on X that pulled in more than 90 million views in under 24 hours, accusing his former employer and OpenAI of 'gambling with our lives' in a race toward self-improving AI. The Guardian's writeup asks why this particular alarm broke through where dozens of prior ones did not.

Part of the answer is who is speaking. Coxon, a Cambridge mathematics graduate who spent about three years doing pretraining work at OpenAI before moving to Anthropic, built the capabilities he is now warning about. 'I resigned from Anthropic today. Neither company is acting responsibly,' he wrote. 'They are racing straight to self-improving superintelligence and gambling with our lives.'

His framing is bleak but not immediate. Coxon has been careful to say current systems are safe for day-to-day use, and directs his concern at trajectory. 'If you have a super advanced intelligence, it could, it will be smart enough to kill us,' he told Time. He described the culture inside frontier labs as one of drift: 'There's this atmosphere of almost resignation where people have accepted that the whole race is happening.'

The link moved through the researcher circles we watch within the same day it went up. Whether that translates into more insider departures, or gets absorbed as another viral moment, is the piece the Guardian is really tracking.