Most health systems already use AI. Far fewer could explain to a regulator how it works.
AI has moved from pilot to production across much of healthcare. Governance hasn’t kept up. A 2025 HFMA report found that 88% of health systems use AI internally, but only 18% have mature governance structures. A Black Book Research survey of 182 hospital leaders, published in November 2025, found similar gaps going into this year’s budgets.
Clinical AI affects patient safety, documentation, coding and payer relationships. That puts it in front of the board and the CEO, not just the CIO. Start with three steps: keep a single inventory of every AI tool in use, name one accountable executive owner, and require vendors to provide documentation you could hand to an auditor.
Executive takeaway: Before approving another AI purchase, ask your team this: "If CMS or a payer asked us tomorrow how this tool reached its recommendation, could we answer within 30 days?" If the answer is no, fund the governance work first.
Evaluate ambient AI scribes through separate measures of note quality, clinician workload, patient trust, and the full cost of implementation.
Read analysis
Cyber resilience is patient safety. See how hospital boards and leaders can govern downtime, sustain essential care, and validate safe recovery.
Read analysis
A narrative review of generative AI governance for hospitals, covering patient safety, evidence, lifecycle controls, equity, accountability, and scale.
Read analysis
