Anthropic Researcher Quits Over AI Safety Fears
Anthropic researcher Jacob Coxon has resigned from the artificial intelligence company, warning that leading AI labs are moving too quickly towards self-improving superintelligence without adequate safeguards. Coxon, who previously worked at OpenAI, accused both companies of “gambling with our lives” in the race to build increasingly powerful AI systems.
Jacob Coxon Quits Anthropic Over AI Risks
Coxon said he had spent the past three years conducting pretraining research at OpenAI and Anthropic before deciding to leave the industry.
In a lengthy statement, he argued that neither company was acting responsibly and warned that future AI systems could become capable of hacking sophisticated systems, rapidly advancing scientific fields and acquiring significant real-world influence.
These are Coxon’s assessments of potential future risks rather than established predictions about how advanced AI will develop.
Researcher Warns of Superintelligence Race
Coxon claimed that some researchers working on frontier AI privately believe highly advanced systems could pose an existential threat to humanity before the end of the decade.
He argued that Anthropic takes such risks seriously but remains caught in a competitive race because slowing development could allow other laboratories to move ahead.
Coxon also cited recent incidents involving autonomous AI agents as warning signs that stronger coordination and safeguards may be needed.
Calls Grow for Stronger AI Safety Measures
The former researcher called for greater cooperation between leading AI companies and suggested that governments may eventually need to consider measures such as temporary restrictions on improvements in model capabilities.
He questioned whether private companies should independently decide when to begin developing potentially superintelligent systems without broader oversight.
Coxon’s resignation adds to an expanding debate within the AI industry over whether rapid advances in model capabilities are outpacing efforts to understand, regulate and safely control increasingly autonomous systems.
