If you thought AI was something you could hand the wheel to and walk away – FDA just pulled over a company for exactly that.
A warning letter issued April 2, 2026 to Purolea Cosmetics Lab lays out something regulated companies in every sector need to understand: AI can support compliance work, but it cannot own it. The letter describes a manufacturer that used AI agents to create drug product specifications, procedures, and master production and control records – and then failed to adequately review those documents to confirm they were accurate and actually compliant with CGMP requirements.
FDA was direct in its response: if AI is used to help with CGMP activities going forward, any output or recommendation from an AI agent must be reviewed and cleared by an authorized human representative of the firm’s quality unit.
That requirement isn’t a technicality. It’s a reminder of where accountability lives – and it’s not in the software.
The Moment That Should Stop Every Regulated Leader Cold
Perhaps the most striking detail in the letter involves process validation. During the inspection, FDA investigators informed the firm that process validation had not been completed before distribution – a legal requirement. The company’s response? They weren’t aware it was required because the AI agent they were using had never told them.
Let that land for a moment.
“The AI didn’t tell us” is not a quality system. It is not a compliance strategy. And it would not hold up in front of an inspector, a customer, a veterinarian, or a board of directors.
In animal health, this matters in a particular way. Our industry runs on clinical trust. The professionals and companies in this space are held to standards by the FDA’s Center for Veterinary Medicine, the USDA, the veterinary community, and animal owners who make decisions based on what we tell them is safe and effective. The accountability structure doesn’t shift because AI helped draft the document.
AI Is Leverage. It Is Not Leadership.
None of this means animal health companies should avoid AI. The lesson here isn’t fear – it’s governance.
AI can be genuinely useful for lean teams navigating complex regulatory environments. It can help organize information, create first drafts, summarize guidance documents, and reduce the administrative burden on technical and regulatory staff. That kind of support has real value, especially in an industry where qualified professionals are stretched thin.
But there’s a difference between using AI to accelerate expert work and using AI to substitute for it.
AI can draft a procedure. A qualified person must determine whether that procedure is scientifically sound, accurately reflects the process, and meets regulatory expectations for your specific products and claims. AI can summarize a regulation. A regulatory affairs professional must determine how that regulation applies to your pipeline, your label, your evidence package. AI can generate documentation. A quality unit must review, approve, and own it.
Speed is one variable in regulated work. It is not the only one.
The Real Risk Isn’t Using AI – It’s Using It Without Guardrails
A well-governed AI process can help strong teams move faster. A poorly governed one can make weak systems look polished while the gaps underneath grow quietly larger.
This is worth saying plainly for animal health companies that are actively integrating AI into marketing, regulatory submissions, technical service content, veterinary education materials, or product launch planning: the output may look complete before the thinking is sound. It may look compliant before a qualified human has verified it is. That is a different kind of risk than most teams are used to managing, because it doesn’t always look like a risk at first.
The strongest organizations won’t be the ones using AI in the most places. They’ll be the ones who have thought clearly about where AI belongs, where it doesn’t, and who is responsible for every decision that reaches a customer, a clinician, or a regulator.
The bot can help draft. The expert still has to sign off.
Is Your Team Clear On Where The Line Is?
If you’re using AI across regulatory, technical, marketing, or quality-adjacent workflows – now is a good time to map where human review is required and who owns the final output. Connect Veterinary Consulting helps animal health companies build communication and compliance strategies that are as rigorous as they are effective. Book a free consultation at the link below to talk through what this means for your team.
