The Physician's Role in AI Procurement: Accountability Before Authority

When an AI tool causes a patient harm, who gets called into the review meeting?

In almost every health system, the answer is the physician. The same physician who was brought into the procurement conversation as a validator after the vendor had already completed four executive demos, built the business case, and received board approval.

That is not a governance design. That is liability assignment after the fact. And it is one of the most expensive mistakes health systems are making in AI right now.

In this episode, Dr. Sarah Matt makes the case for what actual physician leadership in AI procurement looks like — and what the distinction is between being invited to validate a decision already made and having genuine authority over the clinical risk embedded in that decision.

The argument matters practically: health systems that build physician leadership into AI procurement at the authority level, not the validator level, move faster on AI implementation, not slower. Clinical buy-in is not a late-stage change management problem. It is an early-stage governance design decision.

Dr. Matt draws on her experience advising health systems and physician executives who have found themselves accountable for outcomes they did not control, in tools they did not select, for deployments they were brought into after the strategic architecture was already set.

This episode is for physician executives, CMOs, CMIOs, CNOs, and clinical leaders who are currently in, or about to enter, AI procurement conversations — and for the health system CEOs and boards who want to understand why physician leadership in AI keeps underperforming and what the structural fix requires.

What you will take away from this episode:

• The difference between physician validation and physician authority in AI procurement
• Why organizations that build genuine physician authority into AI governance move faster, not slower
• What the accountability gap looks like when something goes wrong, and how to close it before deployment
• The specific organizational design decisions that determine whether physician executives can succeed or are structurally set up to fail

—

Resources & Links:

📖 Get the Book: "The Borderless Healthcare Revolution" is available now on Amazon and major retailers.

💼 Work with Dr. Matt:
Looking for a keynote speaker or strategic advisor?
Visit: drsarahmatt.com

🔗 Connect on Social:
LinkedIn: https://www.linkedin.com/in/sarahmattmd/
YouTube: www.youtube.com/@DrSarahMatt

📧 Subscribe to The Briefing: drsarahmatt.com/newsletter-signup
 
—
Disclaimer:
The views expressed on this podcast are those of Dr. Sarah Matt and her guests. They do not necessarily reflect the official policy or position of any affiliated institutions. This content is for informational and educational purposes only and does not constitute medical advice or a professional consulting relationship.

Transcript

SPEAKER_00 0:01

The best physician leaders I've ever seen in health systems, they're gone in 18 months. Not fired, not failed, gone. Returned to clinical practice, took an advisory role at a startup, joined a consulting firm, or most often went back to doing what they were trained to do before administration, convinced them they could change things from the inside. The standard explanation is usually compensation. And occasionally, you know, that's true. But in my experience, it's actually never the primary reason. The physicians who leave hospital leadership leave because the organizational structure they were recruited into is designed, without anyone intending it, to absolutely neutralize them. So today I'm going to tell you what the organizational design failures are, the three structural pressure points that make physician leadership completely untenable, and what the health systems that are actually keeping their physician leaders are doing differently. So let me start with the org chart, because this is where most of the damage is done before a physician leader even takes their first meeting. The standard health system org chart places the chief medical officer at or near the executive level. On the org chart, this looks like authority, but in practice, it often is not. The CMO in most health systems has influence over clinical policy and direct authority over physician quality and credentialing. However, operational decisions where the budget actually lives, they go through a totally different chain. And this creates an immediate structural problem. The physician leader is accountable for outcomes in domains where they do not control inputs. So they're responsible for care quality, but don't have control over staffing ratios. They're responsible for patient throughput, but don't have control of nursing deployment. They're responsible for technology adoption, but IT reports to someone else. So I mean, this is not a new problem. It's been the architecture of health system physician leadership for decades. What makes it untenable today is that the stakes are higher and the pace is much faster. Digital transformation, AI deployment, workforce restructuring, these are all decisions that require clinical intelligence to execute correctly. But in most health systems, the physician leader's authority stops precisely where these decisions actually begin. So what happens? The physician leader advocates, leadership listens, and then makes the same decision they were going to make anyway. And the physician leader realizes their role is to provide cover, not to drive outcomes. And within 12 to 18 months, they tend to leave. So pressure point number one, decision authority gaps. The gap between what a physician leader is responsible for and what they actually control is the most direct predictor of departure. When a physician leader can't answer the question, what would you have done differently if you had full authority? That's the moment they begin to disengage. I want to be specific about what decision authority means here. It does not mean unilateral control over all clinical decisions, not at all. Health systems are very complex organizations, and shared governance is appropriate in most domains. What it means is for the decisions that require clinical expertise to make correctly, the physician leader has final authority, not advisory input, not a seat at the table, final authority. The health systems that retain physician leaders have done the organizational design work to identify which decisions are genuinely clinical decisions and have built governance structures that give clinical leadership control over those decisions. Not all decisions, the right decisions. Now, if your physician leader's job description says they're responsible for clinical quality and patient safety, and they can't stop a technology deployment that creates a documented patient safety risk, you have a very obvious decision authority gap. And they know it before you do. Pressure point number two. Credibility erosion under administrative override. Here's the dynamic: a physician leader makes a recommendation based on clinical evidence and operational experience. Administrative leadership overrides recommendations, citing financial or operational constraints. This happens once. The physician leader will probably accept it. Trade-offs are real, and good leaders adapt. But then it happens again and again, and eventually, and this is the part health system executives don't fully appreciate, the physician leader's credibility with their clinical colleagues begins to erode. Because the physicians watching know when the CMO makes a recommendation and it gets overridden. The CMO is not actually running clinical strategy. They're managing the interface between administration and the medical staff. So once that credibility is gone, it doesn't come back easily. The physician leader is now seen by clinical staff as administrative, which means their ability to drive clinical change, which is the core of their value prop, is completely compromised. At that point, the rational choice is to leave. Not out of frustration, out of accuracy. The job they were recruited to do is no longer the job they're being asked to do. Pressure point number three: the identity conflict between clinician and executive. This one is less structural and more psychological, but it's equally real. And again, my experience is a little bit different since I switched from clinical practice into technology and then into executive roles. But physicians are really trained in a culture of accountability that is direct and immediate. You make a decision, you see the outcome, you adjust. Hospital administration operates in a very different timeline with far more disfuse feedback loops. So the physician who moves into a leadership role and spends their day in meetings, reviewing spreadsheets, and managing stakeholder relationships is doing work that can often feel profoundly disconnected from the clinical identity that made them actually good at medicine in the first place. And some physicians adapt beautifully and they find purpose in the systems that they work in. But many don't. And the ones who don't, they leave for other reasons. Health systems that are good at retaining physician leaders are intentional about this. They build roles that preserve a meaningful clinical component. Not because the CMO needs to be in clinic to do their job, because the clinical component preserves the identity anchor that keeps the physician connected to why they moved into leadership in the first place. So a physician leader who is entirely out of clinical practice is in most cases a physician leader who is one bad quarter away from attorney to it. For me, I've always kept clinical practice as a small part of my portfolio. Most of the time, it's because I actually like patients. And I feel like if you're going to talk the talk, you need to walk to walk. And that's a key differentiator. So I do charity medicine these days, but only a couple of times a month because I really want to make sure whatever I'm doing, the rest of my business and the rest of my portfolio, is in service of patients. So what do health systems that retain physician leaders actually look like? Consistently, they do three things differently. First, they build decision authority maps before the recruitment conversation. Before hiring a CMO, they identify which decisions are genuinely clinical decisions and are restructured the governance to give clinical leadership actual final authority over those decisions. The job description describes real authority, not titular authority. I mean, come on, you gotta be real here. Second, they create mechanisms for disagreement that do not require the physician leader to lose. There's a difference between the CMO was overruled and the CMO and the CFO reached a different conclusion through a structured process, and here is the rationale. The first completely erodes credibility. The second demonstrates that clinical input is genuinely part of the decision architecture. Third, they negotiate the clinical component at hiring and protect it operationally. The physician leader maintains a minimum clinical practice, specific hours, specific context, whatever it is that makes them feel good about it. And it's treated as a non-negotiable, not as a nice to have that disappears when the quarter gets busy. This is not charity. It's the organizational design that makes physician leadership sustainable and keeps them credible. So if you're currently running in a health system and experiencing physician leader turnover, here's a three-part redesign framework. First, run a decision authority audit. Map every decision the physician leader is accountable for. And then map who currently has final authority over each of those decisions. Where there's a gap, accountability without authority, you've found the source of the problem. Second, build a structured disagreement protocol. When administrative and clinical leadership reach different conclusions, the process for resolving that disagreement needs to be explicit, documented, and visible to the broader medical staff. It shouldn't be closed door overrides. It should be a structured process with a rationale that the medical staff can actually see. And last, protect the clinical component contractually. If you want your physician leader to maintain clinical practice, and if they want to, put it in the employment agreement. Specify hours and context. Make it a performance metric that works both ways. The health system is responsible for protecting that time, not just the physician for maintaining it. Physician leaders don't leave because they can't handle the complexity of health system leadership. They leave because the organizational design they're placed into makes it completely impossible for them to succeed at the job they were recruited to do. So if you fix the design, most of them are going to stay. If you've watched this happen in your organization or seen it go the other way, a physician leader who stayed, rebuilt, and actually changed the system. I want to hear what made the difference. So comment on this. New episodes every Tuesday and subscribe if this was useful. See you next week. It's Dr. Sarah Matt.