On August 18, 2025, the U.S. Food and Drug Administration (FDA) released final guidance titled “Marketing Submission Recommendations for a Predetermined Change Control Plan (PCCP) for Artificial Intelligence-Enabled Device Software Functions.” This guidance provides long-awaited clarity on how manufacturers can manage iterative modifications to AI-driven medical devices while maintaining safety, effectiveness, and regulatory compliance.
Key Takeaways for the Industry
1. Framework for Ongoing AI Modifications
The FDA acknowledges that AI-enabled devices evolve through continuous learning and updates. Instead of requiring a new 510(k) or PMA supplement for each significant update, manufacturers can now seek pre-authorization for specific, planned modifications through a PCCP.
2. Core Components of a PCCP
Every PCCP must include:
– Description of Modifications: What changes will be made to the AI model or device software.
– Modification Protocol: How those changes will be developed, validated, and implemented.
– Impact Assessment: How benefits, risks, and mitigations will be addressed.
3. Flexibility with Accountability
With an authorized PCCP, manufacturers gain the flexibility to implement pre-defined AI updates without resubmitting to the FDA each time. However, deviations from the PCCP may render a device adulterated or misbranded—making compliance and documentation critical.
4. Alignment with Quality Systems and ISO 13485
PCCPs must be supported by a strong Quality Management System (QMS). The FDA has reinforced that compliance with the QSR (soon to be harmonized with ISO 13485:2016) remains essential for AI-driven device changes.
5. Global Implications
As regulatory bodies worldwide align on approaches to AI in medical devices, FDA’s framework will influence both U.S. and international market strategies. Manufacturers that build PCCPs into their submissions now will be well-positioned for global adoption.
How Avendium Can Help
At Avendium, we partner with life science companies to navigate evolving regulatory expectations around digital health and AI. Our team can:
– Develop Regulatory Strategies: Tailor PCCPs for AI-enabled devices to meet FDA, EU MDR/IVDR, and other global standards.
– Design Robust Modification Protocols: Define data management, re-training, performance evaluation, and update procedures that align with FDA expectations.
– Strengthen Quality Systems: Ensure your QMS supports PCCPs by integrating ISO 13485 and FDA QSR requirements into your AI development lifecycle.
– Support Submissions: Prepare and review marketing submissions (510(k), De Novo, PMA) that include PCCPs, ensuring efficiency and compliance.
– Mitigate Risk: Assess the impact of AI modifications on safety, effectiveness, and patient trust while enabling innovation.
The FDA’s new PCCP guidance marks a major step forward in enabling adaptive AI technologies to thrive in regulated healthcare environments. Companies that prepare now will not only accelerate innovation but also maintain patient safety and regulatory confidence.
Source: