Anthropic’s Claude is assisting the company in developing the next, more intelligent iteration of the model, the artificial intelligence lab announced on Thursday.

Claude is responsible for 26% of Anthropic’s model research and development, which, according to the company, means it can complete most of a given task “end-to-end from a high-level prompt” while remaining under human supervision.

The model does not yet operate entirely autonomously.

Nevertheless, approximately 90% of the company’s research and development is conducted in “collaboration” with Claude, which Anthropic said means the model can perform “large chunks of work under close human direction.”

The announcement came as some leading figures in AI, led in large part by Anthropic CEO Dario Amodei, are calling for a slowdown in the technology’s development over safety concerns.

As leaders weigh how quickly to advance AI, “we should do everything possible to minimize the gap between what frontier labs know and what the public knows,” the company said in a blog post.

“This means better measuring the development of AI, reporting on it publicly, and giving society an opportunity to decide how to use this information.”

The company said that models accelerating their own development could make it “more challenging for humans to understand or control these systems.”

It argued that sharing these metrics could lead to a better understanding of how close leading AI labs are to achieving recursive self-improvement, or a model’s ability to autonomously build its successor.

Anthropic also called on other AI developers to regularly share similar metrics, promoting a public methodology so that the numbers could be compared over time and, potentially, across labs.

Anthropic's disclosure left unclear how close the company believes it is to achieving recursive self-improvement, yet the pace at which Claude has increasingly contributed to research and development is notable.

The portion of work Claude “leads,” or does in large part while remaining under human supervision, was none in February.

Six months later, it was leading a quarter of research and development work, reaching that benchmark in August.

The company also disclosed details of the agent oversight measures it has in place, noting that there were approximately 30,000 agents doing research and engineering work as of August.

Oversight measures are crucial for seeing how often agent misbehavior is detected by monitoring systems, the company said.

Anthropic recently committed to setting up external third-party evaluators who will be embedded within the company to monitor safety efforts.

An Anthropic researcher sparked much of the recent debate about AI safety when he resigned last week, warning starkly about the threats the technology poses to humanity.

Amodei, OpenAI CEO Sam Altman, Elon Musk and other tech leaders have since backed the idea of slowing development, but other tech leaders and President Donald Trump have pushed back.