Use of AI Agents leads to the first FDA Warning Letter relating to AI

In the warning letter to Purolea Cosmetics Lab dated 2 April 2026, the FDA criticised the inappropriate use of AI agents in pharmaceutical manufacturing. The company has since ceased its drug production.

What specific issues were criticised?

1. During the FDA inspection, the company told the FDA inspectors that it had used AI agents to assist with compliance with FDA regulations. Specifically, the AI was used to create drug specifications, procedures, and master production or control records that met FDA requirements.

Apparently, these AI-generated documents were used without further review. The FDA expects Purolea, where AI is used as a tool in document creation, to review these generated documents to ensure they are correct and actually comply with CGMP requirements. This lack of review constitutes a violation of 21 CFR 211.22(c).

2. Furthermore, the report criticised the excessive reliance on AI in drug manufacturing operations. For example, FDA inspectors noted a lack of process validation in accordance with 21 CFR 211.100 prior to the distribution of the drugs. The company’s response seems curious: “no such legal requirements were known, as the AI agent used had never indicated that this was necessary”.

Due to the cessation of drug production, these deficiencies had no further consequences. Should production resume and AI be used to support CGMP activities, all results or recommendations from an AI agent must be reviewed and approved by an authorised human representative of the company’s Quality Unit (QU) in accordance with Section 501(a)(2)(B) of the FD&C Act. See also 21 CFR 211.22; 21 CFR 211.100

Conclusion:

The FDA criticised the “blind” use of AI and expects “human-in-the-loop” oversight (= human AI oversight?)