Toggle Daily
Anthropic researcher quits over AI extinction fears
September 15, 2026
Jacob Coxon announced his resignation from Anthropic, saying AI could kill us all by the end of the decade, in a post that shook Silicon Valley, Washington and Beijing.
Anthropic is the AI company behind the Claude model, and AI insiders have issued dire warnings before: every few months for the past several years, a researcher has resigned in a moment of crisis over the same fears.
Key facts
- Fewer than seven days ago, a young researcher at the AI giant used a post on X to say he was leaving, and the message rattled Silicon Valley, Washington DC and Beijing alike.
- Coworkers who remain employed by the company backed up what he said about many AI researchers holding that extinction of humanity is a possible result, and one of them put the odds of that at above 10%.
- The company laid out on Thursday the ways in which criminals and rogue scientists had tried to turn Claude toward building bioweapons.
- His departure came ahead of the point at which equity in the company would have vested, which shows what speaking up cost him personally.
The Guardian reports that Jacob Coxon announced his resignation from Anthropic last week, stating that AI could kill us all by the end of the decade, in a post on X that has shaken Silicon Valley, Washington DC and Beijing. The Wall Street Journal ties the same departure to fears of AI running out of control. Both accounts place the resignation and the reason for it at the center of the event, and they agree on the substance.
The Guardian carries the detail that colleagues still on Anthropic's payroll corroborated his assertion that many AI researchers believe AI could bring about human extinction, with one co worker gauging that outcome as more than 10% likely. It also reports that Coxon resigned before he would have received equity in Anthropic, which shows the personal cost of speaking out, and that on Thursday Anthropic detailed how criminals and rogue scientists had attempted to use Claude to make bioweapons. Amba Kak is quoted saying the fears allowed companies like Anthropic to position themselves immediately as the primary experts in problems they created or are best positioned to prevent.
The resignation, the stated reason, and the corroboration from serving colleagues are settled across both accounts. What Anthropic does in response, and how the warning lands beyond the industry, is not carried.
Kak said: “It allowed for companies like Anthropic and others to position themselves immediately as the primary experts in how to deal with these fears. How to solve problems that, by every account, they have created or are best positioned to prevent.”
Kak
Still developing
The outcome of Jacob Coxon's warning for Anthropic is where the account stands by the time of draft.
How settled the reporting is
Sources
- The Guardian — Why this AI doomsday warning from former Anthropic researcher broke through
Toggle read the full reports of 1 of the 2 outlets counted on this event. The other 1 are counted from their headlines and opening sentences.