AI & Quality Related Warning Letter

Everyone uses AI, and in many cases it is wonderful. However, this warning letter highlights that using AI is not a shortcut for quality and reinforces the “trust but verify (and validate)” adage.

This company makes homeopathic drugs, which, as a regulatory category, have been under FDA scrutiny for a long time. FDA inspected the company, found major GMP and disease claim violations, and additionally cited the use of non-validated AI. Let’s see what we can learn.

This is the fourth FDA warning letter to mention artificial intelligence, but the first to go into detail about best practices and the potential flaws when using this tech in quality systems.

From warning letter. “During the FDA inspection of your drug manufacturing facility, you stated to FDA investigators that you utilized artificial intelligence (AI) agents (b)(4) to help your firm comply with FDA regulations. Specifically, you used AI to create drug product specifications, procedures, and master production or control records to be in compliance with FDA requirements.”

Asa comments: Okay, on face value this could be okay. The use of AI agents can unlock greater efficiency and enable the review of large amounts of data, such as for adverse event trend analysis. However, humans in the loop will always be required because humans, not AI, are the responsible parties. Leveraging AI for critical quality work, such as compiling production records, is dangerous without validation and documentation.

From warning letter. “If you use AI as an aid in document creation, you must review the AI generated documents to ensure they were accurate and actually compliant with CGMP. …Overreliance on artificial intelligence for your drug manufacturing operations was also documented during the inspection. For example, the FDA investigators found that you had not conducted process validation prior to distribution of your drug products, as required under 21 CFR 211.100, and informed you as such. You replied that you were not aware of the legal requirement, as the AI agent you used (b)(4), never told you it was required.”

Asa comments: Leveraging AI for quality work is risky unless the tool and process are validated and documented. Blaming the AI agent for failing to identify a legal requirement is like blaming spellcheck for submitting the wrong word. The tool may assist, but ultimately, the human remains the responsible party.

As many of you know, I am chair of the American Herbal Products Association (AHPA)’s Technology & AI Innovation Committee and founded Apex Compliance®, a compliance technology platform that uses proprietary algorithms, channel integrations, and AI to improve compliance. I love being at the intersection of compliance, tech, and natural products, and I look forward to continuing the discussion on safe and responsible AI use.

Read the warning letter.