The artificial intelligence industry is facing another major internal crisis following the departure of human security researcher Jacob Coxon. In his resignation letter, Coxon accused Anthropic and OpenAI of engaging in a reckless race for self-improving, ultra-powerful artificial intelligence, completely ignoring the risks that exist.
Evan Hubinger, who leads one of Anthropic’s security teams, commented on the situation. He confirmed the concerns of his colleagues, acknowledging that the process of creating uncontrolled artificial intelligence is moving faster than expected and personally estimating the likelihood of a global catastrophe by the end of this decade at more than 10 percent.
At the same time, Hubinger publicly stated that the company currently has no clear plan to ensure the security and consistency of such systems. The scandal has significantly increased pressure on leading artificial intelligence labs, which are speeding up development ahead of preparations for initial public offerings amid rising incidents of autonomous artificial intelligence agents.
