FDA Issues First Warning Letter for Inappropriate AI Use in Drug Manufacturing
Key Insights
The FDA issued its first warning letter to Purolea Cosmetics Lab (search) for excessive reliance on AI systems to generate drug specifications, procedures, and production records without adequate human oversight.
The company used AI to create critical manufacturing documents but failed to review them for accuracy and GMP compliance, leading to violations of quality control requirements.
Legal experts emphasize that the FDA is not against AI use in manufacturing, but requires proper understanding of GMPs and adequate review of AI-generated content.
The US Food and Drug Administration has issued its first warning letter specifically addressing inappropriate use of artificial intelligence in pharmaceutical manufacturing, targeting Purolea Cosmetics Lab (search) for excessive reliance on AI systems without adequate quality control oversight.
The warning letter to the Livonia, Michigan-based company followed a three-day inspection in late October and represents a significant regulatory milestone as AI adoption accelerates across the pharmaceutical industry. The company produces homeopathic drug products, including "Dermveda Extra Strength Shingles Relief (search)" and "Dermveda Extra Strength Ultra Genital Herpes Relief (search)."
Regulatory Violations and AI Misuse
The FDA found that Purolea had relied heavily on AI systems to generate key documents, including drug specifications, procedures and production records, without sufficient human oversight. The agency cited the company's quality control unit under 21 CFR 21.22(c) for its "inappropriate use" of AI in pharmaceutical manufacturing.
Company officials explained to the FDA that they used AI to help the firm adhere to FDA regulations, specifically in the creation of drug product specifications, procedures, and master control records. However, this approach did not satisfy regulatory requirements.
"If AI is used in document creation, the firm must review the AI generated documents to ensure they were accurate and actually compliant with CGMP. Your failure to do so is a violation of 21 CFR 211.22(c)," the FDA stated in the warning letter.
Fundamental GMP Deficiencies
The overarching theme of the warning letter was the company's lack of basic understanding of Good Manufacturing Practices (GMPs) and its reliance on AI to substitute this knowledge. The FDA found that the company's quality control unit failed to ensure that established procedures were followed, that batch records were reviewed prior to product release, and that adequate production and process controls were put in place.
In one particularly concerning example, an investigator noted that the firm had not performed process validation to ensure that its processes were under control. The firm replied that it "was not aware" of the legal requirements because AI did not inform them that this validation was needed.
Additional Manufacturing Issues
Beyond AI-related violations, the inspection revealed multiple other GMP deficiencies. Investigators observed unsanitary conditions at the facility, including the presence of insects, dirt, leaves, and clutter in several areas throughout the facility.
The company also lacked sufficient measures to prevent contamination from both internal and external sources. When the docking bay door was opened, it directly exposed the manufacturing area to the outside environment. The FDA noted that the company failed to conduct "appropriate laboratory testing" to ensure that products were free of objectionable microorganisms.
Legal Expert Perspective
Kalie Richardson, an attorney from Hyman, Phelps & McNamara, emphasized that this warning letter should not be interpreted as FDA opposition to AI use in manufacturing. "I don't get the impression that the FDA has an issue with the use of AI to draft documents. Their issue is that no one reviewed it and it was completely inaccurate," Richardson stated.
Richardson noted that the issues were "so rudimentary" and that "no one has any knowledge of GMPs, and no one knows enough to know what AI is telling you is wrong." She suggested that manufacturers should ideally use a closed system where they know what the inputs are and the source materials are.
Current Status
The FDA noted that the company has stopped producing drugs at the facility. This action represents the first regulatory enforcement specifically targeting AI misuse in pharmaceutical manufacturing, setting an important precedent for how the agency will approach AI oversight in the industry.