Artificial intelligence company Anthropic says its Claude AI model is now playing a significant role in developing the next generation of its technology, marking a major step in the growing use of AI systems to accelerate their own development.
Anthropic said Claude is currently leading about 26% of the company’s model research and development work, according to measurements from independent AI research organization Epoch AI. The figure represents a sharp increase from about 1% in March.
The company said Claude can complete substantial portions of research and development tasks from high-level instructions, although the system still operates under human supervision and is not independently building its successor without human involvement. About 90% of Anthropic’s research and development work now involves some form of collaboration between humans and Claude.
The development has renewed discussion about recursive self-improvement, a concept in which increasingly capable AI systems could eventually help design, train and improve their own successors.
Anthropic itself acknowledges that fully autonomous recursive self-improvement has not yet been achieved and is not inevitable. However, the company says the growing contribution of AI to AI development suggests the possibility could become increasingly important in the coming years.
The company also reported that around 30,000 AI agents were active on its internal research platform in August. Anthropic said the actions of these agents are screened before execution and that human oversight remains part of its development process.
The increasing role of Claude comes amid wider debate within the AI industry about how quickly frontier models should advance and what safeguards are needed as their capabilities grow.
Anthropic CEO Dario Amodei has recently called for leading AI companies to coordinate efforts to slow certain aspects of AI development so that safety measures can keep pace with increasingly capable systems. The company’s latest disclosure is also part of a broader push for greater transparency around how frontier AI labs measure progress.
Anthropic says greater transparency is important because the public currently has limited visibility into how quickly AI systems are becoming capable of performing sophisticated research tasks.
The development does not mean Claude is independently creating a replacement for itself. Instead, it shows how AI is increasingly being integrated into the research process used to develop future AI systems.
As AI models take on larger portions of coding, experimentation and research, the balance between human oversight and machine-led development is becoming one of the industry’s most closely watched questions.
Swifteradio.com
