The US Food and Drug Administration (FDA) has issued a warning letter to Purolea Cosmetics Lab, a Michigan-based drug manufacturer, marking what is believed to be the first enforcement action to explicitly cite AI misuse in pharmaceutical manufacturing as a named cGMP deficiency.
Issued on 2 April 2026, following an inspection of the company's Livonia facility in October 2025, the letter identifies multiple violations of Current Good Manufacturing Practice (cGMP) regulations under 21 CFR parts 210 and 211.
Among them is a dedicated section titled "Inappropriate Use of Artificial Intelligence in Pharmaceutical Manufacturing," which sets this case apart from previous FDA enforcement actions.
According to the warning letter, Purolea used AI agents to generate drug product specifications, standard operating procedures and master production and control records, with the stated aim of supporting FDA compliance.
However, the firm's quality unit (QU) failed to review or verify those AI-generated outputs before they were incorporated into manufacturing operations.
The FDA cited this as a violation of 21 CFR 211.22(c).
When FDA investigators informed the firm that process validation had not been conducted prior to product distribution (which is a foundational cGMP requirement under 21 CFR 211.100), the company responded that it had been unaware of the legal obligation because its AI agent had never flagged it as a requirement.
In its letter, the FDA made clear that this is not a defence.
Instead, the agency's position is that any AI-generated output used in cGMP activities must be reviewed and approved by an authorised human representative of the QU before being entered into the quality system.
Thus, the QU's accountability cannot be transferred to a bot.
The aftermath
Purolea has since ceased drug production, though products remain within expiry in US distribution, prompting further FDA scrutiny of whether they meet established specifications.
The ruling does not necessarily signal FDA opposition to AI adoption in drug manufacturing, but rather reflects the agency's well-established regulatory rule book.
Years of guidance, which include CDER's 2023 discussion paper on AI in drug manufacturing and the January 2026 Guiding Principles of Good AI Practice in Drug Development (developed with the European Medicines Agency), have already set the agency's expectations for manufacturers.
This warning letter suggests enforcement is now following.
For CDMOs, contract testing laboratories and other outsourced manufacturers, the implications are particularly significant — the FDA considers contractors extensions of the manufacturer, meaning AI governance gaps at a contract facility can directly translate into compliance risk for the sponsor.
Quality agreements that do not address AI use, permitted applications, human oversight requirements and audit rights are now considered incomplete.