Is AI the World’s Ultimate Gamble? Anthropic Researcher Quits, Says Models Learning to ‘Kill Us’

An Anthropic researcher who specialized in training new artificial intelligence (AI) models has quit the company and the industry itself, warning that the leading firms developing the technology are racing at a pace that poses existential risks to human civilization. An AI-generated illustration for artificial intelligence. A researcher at Anthropic has resigned on his belief that the companies leading AI are acting irresponsibly. (Image: Shutterstock) Jacob Coxon worked at OpenAI for three years before he departed for Anthropic in July 2026. Coxon said he changed jobs because he thought Anthropic was a better fit since the company had a reputation for prioritizing AI safety. Less than two months later, Coxon wrote on X that Anthropic is no better than OpenAI, and that the AI race to developing the preeminent superintelligence is occurring at an apocalyptic pace. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives,” Coxon wrote. “Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing,” Coxon continued. I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.— Jacob Coxon (@hilbertspaess) September 9, 2026 Gambling on AI Gambling is defined as risking something of value on the uncertain chance of gaining more than staked. AI, if it isn’t already, is expected to change the world for good, leading to breakthroughs in science and medicine, handling repetitive tasks, processing big data, and reducing human error. But whether the benefits outweigh the risks depends on who one asks. Coxon said recent models from OpenAI and Anthropic have demonstrated their abilities to adopt nefarious objectives, posing risks to human life. The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible, but I hear the same people express fear privately. No other human activity poses this level of danger,” Coxon said. Coxon isn’t the first AI researcher to resign over doomsday prophecies. He joins a growing list of former OpenAI and Anthropic researchers and executives who believe AI models could someday refuse human demands and develop a form of free will, allowing the systems to act in ways that prioritize their own survival. AI Odds Prediction markets offer trading on AI-related events, including whether Anthropic will go public through an IPO (traders say there’s a 94% chance the company will before Dec. 31, 2026). OpenAI is the trading front-runner to have the #1 AI model in 2026, with its shares trading at implied odds of 38%. The Polymarket market will settle based on the Dec. 31 rankings on the Arena AI Text Leaderboard, with “Style Control” toggled off. The post Is AI the World’s Ultimate Gamble? Anthropic Researcher Quits, Says Models Learning to ‘Kill Us’ appeared first on Casino.org.