Daily AI signalFeatured
Reel
AI Intelligence Daily
Featured

Anthropic’s biolab made a discovery it’s comparing to Crispr

The discovery involved 950 Claude-powered agents analyzing DNA sequences for 21 hours. The AI identified 200,000 genes for one enzyme type, with 20 selected for further study.

Published 23 September 2026 · ID 2026-09-23-anthropic-s-biolab-made-a-discovery-it-s-comparing-to-crispr
Anthropic’s biolab made a discovery it’s comparing to Crispr

Anthropic claims its AI system, Claude, has autonomously discovered a new enzyme system that resembles the machinery behind Crispr, a gene-editing tool. This marks the first outcome from Anthropic’s newly launched wet lab and serves as an early test of Claude’s potential in scientific research as the company prepares for its public debut.

The discovery was made using 950 Claude-powered software agents that combed through a database of DNA sequences for 21 hours. This process, which would typically take an expert scientist weeks or even months, was completed by the AI in a fraction of the time.

According to the company, the AI system analyzed over 200,000 genes for one type of enzyme and identified 20 of them for further investigation. This work required nearly 1,000 Claude agents and consumed 210 million tokens, highlighting the computational intensity of the task.

The implications of this discovery could be significant for the field of biotechnology. The ability of AI to accelerate scientific research may reduce costs, improve efficiency, and open new avenues for innovation. However, it also raises questions about governance, data security, and the potential for vendor lock-in as companies increasingly rely on AI systems for critical research.

As Anthropic continues to develop its biolab, the broader scientific community will be watching closely. The success of this project could influence how AI is integrated into other areas of research, potentially reshaping the landscape of scientific discovery and innovation.

Sources

Share on X Share on LinkedIn