In a Thursday announcement, Anthropic, the AI lab behind the Claude model, said the system now leads 26% of the company’s research and development work. The company claims Claude can complete most tasks “end‑to‑end from a high‑level prompt” while remaining under human supervision, though it is not yet fully autonomous.
Claude’s expanding role
Anthropic reports that roughly 90% of its R&D is performed in “collaboration” with Claude, meaning the model handles large chunks of work under close human direction. The share of work Claude leads grew from zero in February to a quarter of the effort by August, a rapid increase that the firm says demonstrates the model’s growing capabilities.
Call for industry transparency
Alongside the performance update, Anthropic urged other AI developers to publish comparable safety metrics regularly. In a blog post the company argued that sharing data on how close labs are to achieving recursive self‑improvement – a model’s ability to autonomously build its successor – would give the public a clearer picture of AI progress and help society decide how to use the technology.
Anthropic also highlighted its internal oversight, noting about 30,000 agents were engaged in research and engineering as of August and that the firm plans to embed third‑party evaluators to monitor safety efforts.
Political backdrop
The announcement arrives amid a heated debate over AI policy. Leading AI figures, including Anthropic CEO Dario Amodei, OpenAI chief Sam Altman and entrepreneur Elon Musk, have called for a slowdown to address safety concerns. President Donald Trump, however, has pushed back against calls for a moratorium, emphasizing the importance of maintaining America’s competitive edge in technology.
From a conservative perspective, the President’s stance aligns with the need to protect American innovation while still encouraging responsible development. Anthropic’s own call for transparency does not conflict with a robust, market‑driven approach; rather, it offers a framework for accountability without stifling progress.
What this means for the public
For everyday Americans, the news signals that AI systems are becoming more capable yet remain under human oversight. The push for public metrics could lead to clearer information about how AI is evolving, helping families, businesses, and faith communities understand both the benefits and the risks.
As the debate continues, the Trump administration’s emphasis on preserving technological leadership while addressing safety concerns may shape future regulatory actions. Anthropic’s willingness to share data could set a constructive precedent for industry‑government collaboration.
Original reporting: NBC Connecticut — read the source article.
