Ex-Anthropic researcher warns AI race is reckless and cites leadership flaws
Former Anthropic researcher Jacob Coxon used a Tuesday AMA to criticize the company’s push toward recursive self-improvement and its paranoid stance on China and the US.
Jacob Coxon, who left Anthropic after a brief stint, held an AMA on X where he condemned the company’s “consequentialist philosophy” and its drive toward recursive self-improvement (RSI). He claimed Anthropic’s belief in the inevitability of an AI race has actually accelerated the push for self-improving superintelligence, forcing rivals like OpenAI to abandon projects such as Sora. Coxon warned that many AI engineers believe the technology could be lethal by the end of the decade, describing his statements as anything but a marketing stunt.
He also said Anthropic’s leadership is excessively paranoid about China and the United States, doubting any chance of negotiation. Anthropic CEO Dario Amodei has recently urged frontier AI firms to slow development and focus on safety, a sentiment shared by OpenAI’s Sam Altman, SpaceX’s Elon Musk and DeepMind founder Demis Hassabis. The exchange highlights growing internal dissent and broader industry debate over the speed and safety of AI advancement.
Why it matters
The critique reveals internal doubts about AI race dynamics and safety, influencing public debate and industry policy.
How this story developed
- Sep 9 Anthropic AI researcher resigns, citing existential danger of superintelligent systems
- Sep 9 Evan Hubinger disclosed that some Anthropic staff estimate a greater than 10% chance of human extinction from advanced AI within ten years.
- Sep 10 U.S. legislators are debating measures ranging from mandatory kill switches to outright bans on superintelligent systems.
- Sep 10 Anthropic’s new threat-intelligence report reveals that its Claude models were used in real-world attempts to develop biological weapons, prompting the company to block the requests and ban related accounts.
- Sep 10 Resignation meme spreads across X with satirical adaptations.
- Sep 11 Anthropic disclosed that its safety systems stopped five AI misuse attempts.
- Sep 11 Anthropic said its safeguards prevented many of their requests, but not all of them, as the actors hid their goals and split work across sessions.
- Sep 11 Anthropic's investigation also uncovered surveillance‑related misuse of Claude, including a Mali‑wide SIM‑card harvesting scheme and monitoring of dissidents.
- Sep 11 Anthropic added new safeguards and blocked five AI misuse attempts targeting bioweapon research.
- Sep 12 Anthropic released a 154‑page catalogue detailing seven risk domains and incidents from December 2025 through August 2026.
- Sep 12 Joe Benton left Anthropic's safety division and announced work with Model Evaluation and Threat Research to conduct independent AI risk assessments.
In this story
Related stories
14 in this thread