Ownership in Healthcare AI Governance

One of the issues I keep coming back to in healthcare AI governance is ownership.

Healthcare organizations already have many of the functions needed to govern technology. Legal teams understand the legal and regulatory requirements. Compliance and risk teams have their own responsibilities. Technology and engineering teams build and operate the systems. Clinical leaders understand how the technology may affect care. Vendors may also have an important role, particularly when the organization is using AI that it did not build itself.

The difficulty is that responsibility can become spread across all of them.

This raises a fairly basic question: who ultimately owns the decision? If several groups are responsible for reviewing different parts of an AI system, who is accountable for making sure those pieces come together before it is used?