Anthropic's artificial intelligence model Claude now leads more than a quarter of the company's research and development (R&D) for new AI systems according to a recent Washington Post report.
An internal review revealed that Claude directs about 26% of Anthropic's AI R&D a sharp increase from virtually zero in February. In these tasks Claude executes the majority of the work based on high-level prompts while human researchers provide oversight and final direction. Although Claude is not yet fully autonomous over 90% of the company's R&D now involves the AI model either collaborating with researchers or taking the lead.
Anthropic noted that this rapid integration highlights significant progress toward "recursive self-improvement" a highly anticipated milestone where an AI system can develop more advanced versions of itself with minimal human intervention. However company has not specified how close the industry is to fully reaching this point.
The company also disclosed that by August, approximately 30,000 AI agents were conducting internal research and engineering tasks. To ensure safety, Anthropic has deployed monitoring systems designed to detect and flag potentially problematic AI activities for human review.
Anthropic is now urging other leading AI developers to publish similar operational metrics. The company argues that greater transparency will help governments, researchers and the public better understand and track the rapid pace of AI advancement.
This call for transparency follows recent statements by Anthropic CEO Dario Amodei urged industry leaders to coordinate a slowdown in developing "frontier" AI models to ensure adequate safety protocols are established first.
