Anthropic Researchers Warn AI Could Cause Human Extinction

Two researchers at Anthropic have issued stark warnings that artificial intelligence development could lead to human extinction within years, according to The Japan Times, sparking renewed calls from U.S. lawmakers for regulatory intervention in the AI industry.

Jacob Coxon, a researcher at Anthropic, announced his resignation from the company and stated that "people building AI earnestly believe that it could kill us all by the end of the decade." His warning was publicly endorsed by Evan Hubinger, a scientist at the same company, who confirmed Coxon's assessment was accurate.

The public statements from researchers inside one of the leading AI safety companies have intensified concerns about the pace of AI development. Anthropic, founded in 2021 by former OpenAI executives, has positioned itself as focused on AI safety research, making the warnings from its own staff particularly notable.

Why it matters

When safety researchers at a company explicitly founded to address AI risks resign and warn of existential threats, it signals a potential disconnect between stated safety priorities and actual development practices. The warnings are prompting legislative action at a time when AI capabilities are advancing rapidly and regulatory frameworks remain largely absent. For business leaders deploying AI systems, the episode underscores the importance of understanding not just what AI can do, but what guardrails exist—or don't—around its development.

Growing regulatory pressure

The Japan Times reports that increasing numbers of U.S. lawmakers are now calling for new rules to govern AI systems in response to these warnings. The specific nature of proposed regulations was not detailed in the report, but the timing suggests lawmakers view the researcher warnings as credible enough to warrant intervention.

Coxon's accusation that both OpenAI and Anthropic are "racing toward artificial intelligence advancements without acting responsibly" points to competitive pressures that may be overriding safety considerations. This dynamic—where companies feel compelled to move quickly to maintain market position—is a central concern for those advocating AI regulation.

The warnings come as AI systems have rapidly expanded in capability, with large language models and other technologies moving from research labs into widespread commercial deployment. The gap between technical capability and safety assurance appears to be widening, according to those inside the development process.

These details were first reported by The Japan Times, based on statements from the Anthropic researchers.