Anthropic reports Claude leads 26% of its AI research and development work

Anthropic disclosed that its AI model Claude now handles 26% of the company's research and development tasks, a rapid rise the firm frames as progress toward self-improving AI systems. The claim, reported by the Washington Post, is a corporate assertion with limited independent verification. This matters as it signals accelerating AI self-improvement capabilities, with potential implications for AI safety, labor dynamics, and competitive positioning.