Anthropic researcher quits, warns AI firms are gambling with our future
anthropic training
| Source: Fortune | Original article
An Anthropic researcher has resigned, warning that AI companies are gambling with people's lives, joining a wave of insider departures that raise concerns over the AI race.
Jacob Coxon, a senior pre‑training researcher who has spent the past three years at both Anthropic and OpenAI, announced his resignation from Anthropic on Tuesday. In a post on X, Coxon warned that the industry is “racing straight to self‑improving superintelligence and gambling with our lives,” and called for formal pacing agreements between leading labs to curb the speed of development.
Coxon’s departure adds a new, high‑profile voice to a growing chorus of internal dissent. Earlier this month, Anthropic researchers warned that there is more than a 10 % chance that advanced AI could pose an existential threat, and the company recently declined to submit its Mythos 5.1 model for pre‑release testing in the UK, fueling fears of protectionist drift. Coxon’s critique underscores the same concern: that competitive pressure is outweighing safety considerations, potentially accelerating the arrival of systems that can improve themselves without clear oversight.
The resignation matters because it signals that even senior engineers, who have helped shape the core models, view the current trajectory as reckless. Their public statements increase pressure on companies to adopt transparent safety protocols and give regulators fresh ammunition for stricter oversight. Industry observers will be watching how Anthropic and OpenAI respond—whether they will endorse pacing accords, accelerate internal safety audits, or defend the pace of their research.
Going forward, the AI community is likely to see intensified debate over “pacing agreements,” possible legislative action in Europe and the United States, and perhaps further insider departures if safety concerns remain unaddressed. As we reported on 9 September, Anthropic’s own researchers have already flagged a non‑trivial risk of AI‑driven catastrophe; Coxon’s exit brings that warning from the lab bench to the public arena.
Sources
Back to AIPULSEN