New AI Code of Conduct and Oversight Challenges in US Healthcare
June 11, 2025
The National Academy of Medicine released a comprehensive code of conduct for healthcare AI, highlighting governance, oversight, and equity issues.
Key Details
- The 206-page AI code of conduct aims to align healthcare AI stakeholders around governance and ethical best practices.
- The framework includes a 'Tight-Loose-Tight' model: national alignment, local experimentation, then rigorous evaluation and reporting.
- The code is not legally binding but provides essential guideposts for developers, clinicians, and patients.
- Ongoing concerns are raised about inadequacy of FDA oversight and regulatory resources, especially as AI models are updated in real-world use.
- Recent studies show generative AI can boost radiologist productivity by up to 40% and reduce prostate MRI workload by 20%.
Why It Matters
A robust code of conduct is critical for safe and equitable AI adoption in healthcare, including radiology. Real-world oversight gaps and demonstrated productivity gains mean radiology professionals need to engage with these evolving standards and regulatory considerations.