Anthropic acquires AI biotech startup Coefficient Bio in USD 400 million deal
Follow UsAnthropic has acquired Coefficient Bio in a $400 million deal, signalling a stronger push into healthcare AI despite recent security concerns.
The move aligns with Anthropic’s strategy to build industry-specific AI tools for life sciences and medical workflows.
The deal comes shortly after a Claude source code leak, raising fresh security concerns around the company.
After source code leak controversy, Anthropic has reportedly acquired AI biotech startup Coefficient Bio in a deal valued at around $400 million, as per report by The Information. This comes amid the AI companies have been expanding into AI industry specific AI solutions, specifically in healthcare and life sciences.
SurveyAccording to the report, the small but specialised team at Coefficient Bio will join Anthropic’s healthcare division following the acquisition. Despite its recent formation, the startup has created a platform to assist artificial intelligence in managing complex biotech workflows such as drug discovery, research planning and regulatory strategy development.
The acquisition comes after Anthropic announced plans to expand its AI capabilities beyond general-purpose tools into domain-specific applications. In recent months, the company has been actively working to improve its AI assistant Claude by adding features tailored to scientific and medical use cases. These include integrations with platforms like Benchling and BioRender, as well as tools for drafting clinical protocols and regulatory documents.
The company has also announced collaborations with major pharmaceutical and research companies such as Sanofi, Novo Nordisk, Genmab, AbbVie, the Allen Institute, and HHMI. Earlier this year, the company added features that enable healthcare professionals and insurers to use its AI systems in accordance with medical data regulations.
Meanwhile, Anthropic recently accidentally leaked the Claude Code source code through a misconfigured npm package. The leak revealed 512,000 lines of code, including hidden features such as KAIROS (autonomous mode) and Undercover Mode. Hackers are now using leaked versions on GitHub to spread malware and information stealers. If you updated the tool during the leak window, look for malicious axios dependencies (1.14.1/0.30.4) to avoid security issues.
