AI and Digital Oversight in Pharma Supply Chains: PDA Regulatory Conference Insights

Scientist Managing Prescription Drug Supply Chain | Image Credit: © leowolfert - stock.adobe.com

At the Parenteral Drug Association (PDA) Regulatory Conference 2025 in Washington, DC, a session on Sept. 9, 2025 titled “Quality Oversight in a Modern Supply Chain” explored how artificial intelligence (AI), digital governance, and data integrity are shaping quality assurance in pharmaceutical manufacturing. Presentations by Charles Gibbons, director of Data Integrity & Data Governance at Lachman Consultants, and Michael Grischeau, director of Data Analytics and Management Review at AbbVie, highlighted both regulatory expectations and practical use cases.

How are regulators guiding AI use in pharmaceutical laboratories?

Gibbons began his talk, “Beyond the Bench: AI-Powered Oversight for Chem and Micro Labs,” by polling the audience on their AI readiness (1). Responses suggested that while few companies have an approved AI strategy for quality control (QC), most have established governance policies.

“Governance is key,” Gibbons emphasized, noting that “if you don’t have good data in there, you won’t have good data coming out.”

Gibbons also underscored the alignment between United States and European regulators.

“Earlier this year FDA released the draft guidance [on ‘the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products’ (2)], which is a great guidance for everyone—very descriptive and it focuses very much on context of use,” he said.

Parallel updates from the European Medicines Agency, including Annex 11 revisions and the new Annex 22, point toward global convergence on AI oversight, Gibbons said (3).

For laboratories, Gibbons outlined AI applications ranging from faster root cause analysis to real-time anomaly detection. He described how natural language processing can accelerate investigations and how virtual witnessing might “improve reliability because the machine is watching, monitoring, and can also allow for early detection of anomalies.”

However, he said explainability remains important.

“If you don’t understand that, can’t trust that model, or explain that model, it’s going to cause a huge problem,” Gibbons said.

And for the most part, Gibbons said generative AI remains off the table for regulated QC activities, for now.

“At this point, I don’t believe there’s any situation where you’re going to let AI loose and just make decisions on its own,” he concluded. “That human supervision piece is key.”

What’s needed to build an AI-ready supply chain?

Grischeau’s presentation, “Guardians of Quality: Digital Tools and AI in the Era of Complex Supply Networks,” extended the conversation from laboratory operations to enterprise-wide supply chains (4).

“Data is the lifeblood of any technology-driven initiative,” Grischeau told the audience, saying that without good data governance, even the most advanced AI systems will fail—but technology alone is not the barrier. “In my situation, usually the data [or] technology isn’t the barrier; a lot of it is process.”

Grischeau urged companies to invest in process harmonization and standardization. He also pointed to the importance of organizational readiness, including education and change management.

“Making sure that we’re enabling teams across the enterprise and really understanding what is responsible AI and responsible use of AI” is as important as the tools themselves, Grischeau argued.

What are the most practical AI use cases across the pharma supply chain?

Grischeau presented three use cases that illustrate how digital tools are already supporting oversight in complex supply networks.

• Supplier quality monitoring. AbbVie automated its contract manufacturing organization (CMO) scorecard, which transformed a manual, spreadsheet-driven process into a qualified application that was able to be refreshed nightly. This automation “eliminated the need for manual data processing,” Grischeau said, while ensuring transparency with external partners.
• Manufacturing performance. By digitizing certificates of analysis, the company gained the ability to automate trending and performance monitoring across internal and external manufacturing sites. Human verification remains part of the workflow, but efficiencies and onboarding speed improved dramatically.
• Post-market surveillance. AbbVie has deployed an AI-driven complaint coding system that uses generative AI prompts to classify and risk-rank patient complaints. “We’ve moved from minutes to seconds in coding at intake,” Grischeau said, adding that this has enabled faster escalation of high-priority issues and closer collaboration with CMOs.

Across these examples, Grischeau stressed that compliance and monitoring plans remain central.

“You have to implement a solution, you have to monitor the AI output, you have to make sure that you’re compliant, and you have to make sure that you’re driving continuous improvement,” he said.

What does AI mean for drug discovery, development, and manufacturing?

Collectively, the presentations in the “Quality Oversight in a Modern Supply Chain” session illustrated a dual challenge for the pharmaceutical industry: developing AI strategies that improve efficiency and compliance while maintaining regulatory alignment and human oversight.

For drug discovery and early development, where data integrity is often less formally regulated, these insights reinforce the value of clean, well-governed datasets that can later support downstream manufacturing and QC. In drug development and manufacturing, where reproducibility and regulatory scrutiny are paramount, AI offers opportunities to accelerate investigations, monitor supplier performance, and improve responsiveness to market feedback.

The session also highlighted the convergence of technical and organizational factors. AI adoption is not just about algorithms but also process redesign, workforce readiness, and transparent collaboration across global supply chains.

As the presenters made clear, AI is already reshaping quality oversight in pharma. But whether in the laboratory or across the supply chain, its success depends on governance, explainability, and the irreplaceable role of human judgment.

References

1. Gibbons, C. Beyond the Bench: AI-Powered Oversight for Chem and Micro Labs. Presentation at the PDA Regulatory Conference, Washington, DC, Sept. 9, 2025.
2. FDA. Draft Guidance for Industry and Other Interested Parties, Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products (CVM/OII/OCE/CBER/CDRH/CDER/OC/OCMO/OCP, January 2025).
3. Barton, C. Stakeholders’ Consultation on GMP Annex 11, Chapter 4, and New Annex 22. PharmTech.com, Aug. 22, 2025.
4. Grischeau, M. Guardians of Quality: Digital Tools and AI in the Era of Complex Supply Networks. Presentation at the PDA Regulatory Conference, Washington, DC, Sept. 9, 2025.