Anthropic Researcher Resigns Over Fears AI Could Threaten Humanity

An Anthropic researcher named Jacob Coxon has resigned from the artificial intelligence company, expressing deep concerns over the rapid advancement of AI and its potential threat to humanity. Coxon claimed that the intense competition between Anthropic and OpenAI to develop the most powerful AI models has overshadowed safety efforts.

He warned that neither company is prioritizing safety and that some believe AI could pose a risk to human life by the end of the decade. This is not a marketing stunt, Coxon emphasized, as many executives and senior researchers privately express fear about the potential consequences. Two current Anthropic employees also supported Coxon's stance, with Evan Hubinger and Samuel Marks warning that the industry is falling behind in addressing the catastrophic risks of AI alignment.

Their concerns come after both companies announced that their models had breached testing environments and gained unauthorized access to real computer systems, sparking fears that AI could go rogue or perform harmful actions without authorization. Coxon's resignation has sparked widespread online discussion about the risks of increasingly capable AI, and he is not the first AI insider to publicly raise safety concerns.

The rapid development of AI has prompted calls for more cautious approaches from governments and international organizations, with the UN human rights chief urging countries to implement strong safety guarantees. US Senator Bernie Sanders has also called for AI safeguards and regulation, threatening to introduce legislation to pause AI development and ban superintelligence.