OpenAI Research VP Criticizes Anthropic's Perceived Stance That Only They Can Build AI

According to Aidan Clark, OpenAI's research vice president for training, on X platform recently, he expressed criticism of what he has heard from Anthropic colleagues: a belief that only Anthropic is qualified to be trusted with building AI. Clark argued that having multiple organizations developing AGI (artificial general intelligence) is beneficial, while going it alone makes it nearly impossible to find the right path.

Anthropicemployees responded to clarify their positions. One user denied the characterization, stating "that's not our view," while another distinguished between Anthropic not trusting OpenAI specifically versus believing no one else should build AI. Clark acknowledged the distinction, admitting he may have misheard the original claim, but noted that distrust of OpenAI alone is "not much better."