Lessons from GMP By the Sea – Inside Perspectives by Jenn Catour

In a year that’s been marked by significant changes at the U.S. Food and Drug Administration (FDA), the effects were felt at this year’s GMP by the Sea Conference held August 10-11, 2025, in Cambridge, MD. Here are my takeaways – and lessons – from the event applying to Good Manufacturing Practices (GMP) in our current climate. 

Conference Takeaways and Trends 

• Deb Autor, CEO at Healthcare Innovation Catalysts and President of the FDA Alumni Association, summarized the changes at FDA. Changes are happening rapidly! A big shift she highlighted was the return to “Podium Policy.” Examples such as the FDA Direct Podcast covering a wide range of topics and policy pronouncements in media appearances, journal articles, and op-ed pages are creating a regulatory gray zone the agency was criticized for 25 years ago. 

• While current FDA participation was limited, as announced ahead of the meeting, Michael Davis, Deputy Director at CDER, was able to participate remotely and presented a Center update including the Center for Drug Evaluation and Research (CDER)’s Mission and Programs, the current GMP compliance landscape, enhanced inspection and oversight, innovation in GMP assessment, compounding, and drug shortages. Deputy Director Davis emphasized he was excited to be back at CDER, and their 2026 priorities would be focused on efficiency, giving their staff the support to do the work, and the appropriate use of AI (i.e. AI can’t replace human assessments). 

• Hot topics this year were related to changes at the FDA, GMP applications of AI, remote audit tools, Quality Maturity Management, and the always popular “Ask the (former) Regulators” Q&A session to round out the conference. 

Artificial Intelligence – Lessons from Industry Leaders  

• I appreciated hearing about real-world applications of AI in use within GMP systems. Jeff Orlov, Corporate Quality – QMS, Senior Director from Bristol Myers Squibb, presented “Implementation of Quality AI to Support Speed to Patients” and BMS’s “Deviation Draft Assistant” that was developed. BMS took a risk-based approach to their first foray into the use of AI, focusing on “Low Risk - No Impact Deviations” that were not only high in volume across the global network, but also time-consuming to complete. In the time since implementation, BMS has realized a 35% efficiency gain in deviation open to close time, and a 45% gain in % Right First Time (RFT). Tammy Hanley, Head of Performance, Insights & Transformation at Sanofi, shared their new tool “GenAIR” – GenAI powered reports, developed to support the Product Quality Review (PQR) process. PQR processes are a highly regulated activity with intense scrutiny by regulators, and typically very burdensome and time-consuming as a manual data collection exercise. Sanofi achieved major efficiency gains with GenAIR, dropped PQR development time, including data collection, content drafting and formatting, data analysis, review and closure, and final report review and approval from 130 hours per PQR down to 15-45 hours. This result was a 65-85% efficiency gain. In both organizations, large language models were used and these tools enhanced quality and compliance and improved operational efficiency. 
• I attended two afternoon workshops continuing the AI theme. 
  

•  “The Road to AI: Why and How Data Governance Should Pave the Way” was hosted by presenters who emphasized companies should get familiar with the two FDA draft guidance documents that came out in January 2025. We worked in small groups to develop an AI project or tool, creating our own “roadmap” to address data management and address readiness for implementation and risk to the business. 

• “AI Integration – Tools Supporting Document Writing” walked participants through several real-world case studies, the pros and cons of Google Gemini vs. Ghat GPT, and watchouts for introducing and using new AI tools that come on the market claiming to be 21 CFR Compliant.

• These presentations and workshops helped to demystify AI and companies, proposing to lean into low-risk, high impact settings, while applying appropriate validation strategies. As Deb Autor asked in the Q&A session, “How do we unleash AI? How do we get to fully implement AI for GMP activities?” As she said, the potential is there, but we haven’t figured out how to have these regulatory conversations yet. 

Top Innovations and Game-Changing Ideas 

• Ranjith Ramakrishnan, Head of Quality Assurance - Biologics from Syngene, shared his company’s experience hosting remote regulatory inspections, and their use of technology.  In the early days of COVID-19, they utilized GoPro cameras for the site visits/facility tour portion of the remote inspections. Document review and interaction with the inspectors were managed via video cameras attached to laptops. There were many challenges such as device portability, lack of clear images, and poor video quality. These challenges compelled the site to evaluate new technologies such as Gimbal (mobile device) and Head Mountable Devices (HMD) or wearables. The technology that Syngene ultimately deployed was a wearable – the Navigator 520 Assisted Reality (aR) glasses from realwear. From their real-world experience, Syngene now has an “any time” inspection ready status for labs and manufacturing areas, but as an added benefit, they’re able to have real time collaboration during Quality investigations, can more easily collaborate with teams for technical support or troubleshooting with equipment, and can record activities for more robust technical training of employees. 
• Remote assessments are here to stay, and the use of new technology can help companies have more meaningful engagement with regulators, leading to successful inspection outcomes. 

The Future of Our Industry 

Multiple presenters emphasized industry should take advantage of the dynamic regulatory environment and embrace FDA’s invitations to engage in non-traditional forums such as listening tours, sharing sessions and voluntary participation in ongoing programs such as the Knowledge-Aided Assessment and Structured Application (KASA) and Quality Management Maturity (QMM) Initiative.  

Industry should consider these options as a pathway to engage with FDA to gain the appropriate advice early on. A company’s participation in these initiatives shows they’re working proactively on continuous improvement, providing the potential for regulatory flexibility that can impact future Pre-Approval Inspections (PAIs) and/or routine inspections. I fully agree. 

To continue the quality and compliance conversation or discuss any of these lessons from the event, contact us. 

References:  

1. “Considerations for the the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products”: Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products | FDA 

1. “Artificial Intelligence – Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations”: Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations | FDA