Anthropic's Latest Reveal Shows AI Is Moving Toward a Milestone

Anthropic says its Claude chatbot is now a key architect of its own successors, the Washington Post reports. The AI company reported Thursday that Claude is "leading" about 26% of its research and development tasks—with humans largely supervising—up from 0% in February. The milestone edges closer to what experts call "recursive self-improvement," in which AI systems meaningfully advance their own capabilities without requiring input from humans.

Anthropic stressed that Claude does not yet operate fully on its own and said it has about 30,000 AI "agents" working on research and engineering, with roughly 6% of R&D computing power devoted to safety. The announcement comes amid growing calls, including from within the AI sector itself, for more federal guardrails around AI development. "Models accelerating their own development could make it more challenging for humans to understand or control these systems," Anthropic said in its statement. "It is therefore important to share these metrics to understand how close the world is to reaching recursive self improvement."