Former Anthropic Researcher Warns AI Could Cause Human Extinction by 2030

Jacob Coxon, a former researcher at the AI safety firm Anthropic, has resigned and issued a public warning that artificial intelligence could lead to human extinction by 2030. The announcement, delivered via X, suggests that the internal culture of AI development is far more fearful than public corporate messaging indicates.

Coxon asserted that many of the developers and executives currently building these systems earnestly believe that AI could kill all humans by the end of the decade. He emphasized that these concerns are not a "marketing stunt," but rather a genuine fear shared among those with the deepest technical understanding of the technology. According to Coxon, while senior researchers and executives may use cautious or "sensible" phrasing when speaking to the press, they express much deeper anxieties in private.

The Safety Paradox

Anthropic was founded by former OpenAI executives with a specific mandate to prioritize AI safety and the creation of "steerable" systems. Unlike many of its competitors, the company has positioned itself as a safeguard against the uncontrolled growth of artificial general intelligence. However, Coxon's departure highlights a growing tension within the industry. His warnings align with a broader trend of "doomer" narratives, where insiders suggest that the public-facing optimism of the AI boom masks a profound internal dread regarding the potential for superintelligent systems to escape human control.

Systemic Implications

The significance of this warning lies in its origin. Because Coxon comes from a company dedicated specifically to the science of AI safety, his claims suggest that the very safeguards designed to prevent catastrophe may be viewed as insufficient by the people building them. If the industry's primary safety architects believe that extinction is a plausible outcome within a six-year window, it implies a systemic risk that current regulatory frameworks and technical alignment efforts are failing to address.

Looking Ahead

As the race for more powerful models accelerates, the industry faces increasing scrutiny over whether safety can keep pace with capability. While Anthropic continues to promote its commitment to steerable AI, Coxon's resignation serves as a stark reminder of the existential stakes involved. Observers will now be watching to see if other insiders come forward with similar warnings or if the industry can produce verifiable evidence that the risks of extinction are being effectively mitigated.