AI researcher quits Anthropic, warns of safety risks
AI safety concerns grow as an Anthropic researcher warns of rushed development See why leaders want stronger rules, better alignment, and a slower path forward
An artificial intelligence researcher has resigned from Anthropic and criticized both Anthropic and OpenAI for moving too quickly on advanced AI systems. The researcher, Jacob Coxon, said in a post on X that he left because he believes the companies are not handling the technology responsibly.
Coxon said some people building AI think the systems could pose extreme risks within the next decade. His post drew wide attention online and added to an ongoing debate in Silicon Valley about whether advanced AI can be developed safely and controlled effectively.
Anthropic alignment lead Evan Hubinger later echoed some of those concerns, saying the industry does not yet have a clear plan for solving alignment problems for superintelligent systems. The article also notes that lawmakers and researchers are increasingly discussing how to slow development, improve safety, and create rules for advanced AI.