AI-ready department chairs can help turn interest in artificial intelligence into useful progress: better-supported faculty, thoughtfully redesigned workflows, and new opportunities for teaching and research. The practical work begins with choosing priorities, bringing the right people together, and establishing how the department will learn from experience.
For deans and search committees, this adds a specific dimension to leadership assessment. Alongside clinical credibility, scientific accomplishment, and educational leadership, consider how candidates would help faculty explore AI and evaluate its contribution to the academic mission.
The timing matters. Among the 1,342 physicians answering the adoption question in the AMA's March 2026 survey, 72% reported incorporating at least one named AI use case into their practice. These are self-reported responses, not independently verified deployments or evidence of improved outcomes.
A federal analysis of 2024 hospital survey data offers another useful signal: 74% of hospitals in its predictive-AI governance analysis reported that multiple entities were accountable for evaluation. Shared oversight can be appropriate. A chair who understands who recommends, who decides, and who evaluates results can help promising proposals move through the institution with clearer expectations.
Define the mandate before opening the search
Begin with the decisions the incoming leader will own. In a department chair search, that may include selecting specialty priorities, supporting faculty development, and coordinating trainee supervision. For a vice dean, the relevant responsibilities may span an educational or research portfolio. Align the mandate with the dean and health system leadership before assessing candidates.
Write those responsibilities into the position profile. Identify the relationships with information technology, clinical informatics, quality and safety, research leadership, legal counsel, and faculty governance. Specify which decisions the role can make and which require joint approval.
Give the chair or vice dean a clear connection to enterprise decision-makers. The NIST AI Risk Management Framework's governance provisions explicitly place responsibility for AI risk decisions with executive leadership. The framework is voluntary guidance, not an executive job description or a legal mandate.
Assess five capabilities in AI-ready department chairs
1 Turning collaboration into decisions
Ask: “Walk us through an initiative that required clinical, technology, and academic stakeholders to work together. How did you reach a decision and move it forward?”
Look for a candidate who can describe a functioning process: an accountable sponsor, meaningful input from affected groups, a way to resolve disagreements, and a documented decision. Ask how faculty and staff helped shape the proposal and how the candidate kept their participation productive.
Reference checks should establish the candidate's actual role, including how they secured the authority and resources needed to carry out the decision.
2 Building a sound foundation for expansion
Ask: “What would you need to know before expanding an AI tool beyond its initial setting?”
A strong answer should distinguish a vendor's performance claims from evidence in the intended workflow and population. Ask how the candidate would involve appropriate experts in examining errors, subgroup performance, staff workload, and unintended consequences.
The Duke-Margolis white paper on health system AI governance, published in October 2024, describes review processes spanning performance, privacy, safety, and integration. It also documents different institutional approaches. Search committees should reward judgment about the level of review a use requires, rather than demand one universal approval process.
3 Connecting investment to useful results
Ask: “How would you decide whether this tool deserves continued funding after its pilot?”
Look for a clear baseline, a defined problem, and a cost model that includes implementation, integration, training, evaluation, and ongoing support. The candidate should explain how benefits would be measured and who would verify them.
Time saved does not automatically become additional patient capacity or financial savings. Ask what operational changes would be necessary to realize either outcome. A disciplined leader should also be comfortable recommending a workflow change without purchasing another tool.
4 Developing faculty and learners
Ask: “How would your approach differ for an established clinician, a resident learning clinical reasoning, and a faculty member using AI in research?”
The AAMC's principles for responsible AI use in medical education, updated in July 2025, emphasize educator development, interdisciplinary curriculum design, data privacy, and evaluation in the intended setting.
For hiring purposes, ask candidates to translate those principles into faculty time, supervision practices, and assessment methods. Probe how they would coordinate with research leadership when a tool developed for investigation is proposed for routine operations. The candidate should identify which experts and institutional review processes need to participate, without treating research and clinical implementation as interchangeable.
5 Creating a durable learning process
Ask: “How would you keep learning after implementation, and what evidence would lead you to expand, revise, or pause an initiative?”
Look for attention to faculty feedback, incident reporting, changes in the tool or workflow, and responsibilities for follow-through. Ask about an actual initiative where experience changed the original plan. A thoughtful account of adjustment shows how a leader can help a department improve its approach over time.
Make the assessment consistent across finalists
Give finalists the same nonconfidential description of an institutional decision under consideration. Ask each to outline the evidence needed, stakeholders involved, decision authority, budget implications, and conditions for reconsideration. Use the same evaluation criteria and reasonable preparation expectations for everyone.
Score the quality of reasoning, the clarity of ownership, and the candidate's ability to recognize missing information. Avoid treating technical vocabulary or access to a prestigious former institution as a substitute for demonstrated judgment. Verify material claims through references and permissible examples, without requesting confidential documents from prior employers.
Carry the search mandate into onboarding
Before appointment, agree on a realistic first-100-day agenda. Depending on the role and existing infrastructure, priorities might include meeting faculty champions, reviewing current AI uses, agreeing on shared evaluation priorities, and proposing a funded faculty-development plan.
These are early management deliverables, not promises to demonstrate clinical or financial returns within 100 days. The appointment should establish who will assess progress and how decisions will reach the appropriate governing bodies.
For institutions preparing a department chair search or vice dean search, Academic Med can help frame the role around the decisions and opportunities ahead. A well-defined mandate gives the search committee a stronger basis for evaluating candidates and gives the incoming leader a clear foundation for advancing the academic mission.