AEGIS: An Operational Infrastructure for Post-Market Governance of Adaptive Medical AI Under US and EU Regulatory Frameworks
CONCEPTNature publishes research on AEGIS, an operational infrastructure framework for governing adaptive AI medical devices that can change their behavior after deployment. AEGIS addresses the challenge of maintaining safety and effectiveness for AI systems that evolve and learn in clinical settings. Adaptive AI systems pose novel governance challenges because they may perform differently over time or in different clinical contexts than they did during initial validation. AEGIS proposes systematic approaches to ongoing monitoring, validation, and governance of adaptive AI systems under both US FDA and EU regulatory frameworks. This research contributes to the emerging field of adaptive AI governance, addressing a critical gap in current regulatory frameworks that assume medical devices remain stable after approval. ---
Subscribe to read full wiki analysis
Full wiki content — including source provenance, contradiction tracking, related pages, and analyst notes — is available to Subscriber members.
View Plans & SubscribeNew users get a 7-day trial — no credit card required.