How Healthcare Leaders Can Enable Human-Centered AI

Artificial intelligence (AI) is no longer a futuristic idea for hospitals and clinics. It’s here now. Healthcare is increasingly using AI in real ways—from easing clinician paperwork to improving patient care. But technology alone won’t improve healthcare. Leaders must drive adoption in a human-centred way. This means focusing on people first—staff, patients, and the entire care experience.

At the CHIME Fall Forum in 2025, healthcare IT leaders shared real stories of AI in action. They discussed the successes, challenges, and what it takes to adopt AI responsibly. Leaders from Harvard Medical School also explain why AI adoption requires strategy, training, and ethical oversight. Together, these insights show how healthcare IT leaders can lead AI adoption that works for people—not just technology.

Why Human-Centred AI Matters in Healthcare

AI tools are becoming more common in healthcare. They can help with paperwork, clinical decisions, and workflows. But AI must be used in ways that benefit patients and staff. Technology should support care, not slow it down. It should support it.

Human-centred AI means:

AI improves experiences for clinicians and patients.
Teams trust the technology and know how to use it.
AI tools are safe, accurate, and easy to work with.

Dr Ted James from Harvard says that AI should augment, not replace, human work. That means AI should help people do their jobs better, not make them obsolete. Leaders need to guide teams to see AI as a partner, not a threat.

Start with a Clear Purpose, Not Just Technology

One mistake organisations make is adopting AI without a clear goal. Leaders should always begin with a problem to solve, not a specific tool to buy.

At the CHIME25 forum, Chris Harper from the University of Kansas Health System shared how they used an ambient dictation tool to reduce clinician documentation work. They didn’t start with the tool. They began with the problem: heavy documentation burden. After piloting with 25 providers, they saw significant time savings. Physicians could see more patients, reduce burnout, and improve care.

What this shows is simple: pick use cases that matter. Leaders must work with clinical and administrative teams to find priorities where AI can add real value. Start small, measure outcomes, and then scale based on results.

Build Trust Through Pilot Projects

AI tools can be powerful—but they won’t succeed if people don’t trust them. Leaders must create opportunities for teams to see AI work firsthand.

Many organisations use pilot testing to build trust. Start with a small group of users. Test the tool. Gather feedback. Then improve the solution. This “fail fast, learn fast” approach helps teams feel comfortable experimenting without fear.

At Corewell Health, leaders are piloting several AI solutions. They don’t wait for perfection. Instead, they try tools quickly, learn what works, and adapt. This helps staff see success early and builds trust over time.

Create a Culture of Learning and Innovation

AI adoption is not only about technology. It’s about people and culture. Leaders must help their workforce feel confident and skilled when using AI. That means training, education, and support.

Dr Ted James highlights that successful AI adoption depends more on culture than technology itself. Leaders should encourage experimentation and continuous learning. This includes mentoring, hands-on training in AI tools, and encouraging staff to ask questions.

This culture helps teams adapt to new roles where human judgment and AI work together. Training reduces fear and increases confidence in using AI in daily workflows.

Integrate AI with Real Workflows

AI tools have the most impact when they fit naturally into existing workflows. If a tool interrupts care, people won’t use it.

At San Joaquin General Hospital, leaders learned this lesson the hard way. They implemented an AI scribe for clinical documentation. But doctors in different departments had different needs. Emergency doctors loved the notes. Inpatient teams found them too detailed. The tool wasn’t customised for each workflow, which caused frustration.

Healthcare IT leaders must involve stakeholders from all departments early in the process. This ensures that tools work well for every team that uses them. Leaders shouldn’t force a technology on people—tools must align with daily work.

Measure Success: Define Metrics Up Front

Clear goals and performance measures must guide AI adoption. Metrics help leaders understand whether tools are making a positive impact.
Good metrics include:

Time saved on tasks
Reduction in errors
Impact on clinician workload
Improvements in patient outcomes
Cost savings over time

At the University of Kansas, leaders continued to measure outcomes even after scaling their solution. This ensured that the tool stayed aligned with expectations and continued to deliver value.

Ensure Governance, Safety, and Ethics

AI involves sensitive data. Leaders must ensure that tools are safe, ethical, and compliant with privacy laws. HealthTech speakers highlighted the need for strong AI governance. AI tools must be evaluated for:

Security and data protection
Accuracy and bias
Transparency of algorithms
Alignment with health regulations (like HIPAA)

Tracey Touma from the Cleveland Clinic stressed that governance is a human challenge, not just a technical one. Teams must establish policies and committees to evaluate AI tools continuously. This helps reduce risks like data breaches, bias, and unsafe outcomes.

Include Everyone: Build Cross-Functional Teams

AI adoption isn’t only an IT project. It requires collaboration between IT, clinicians, administrators, and business leaders.

HealthTech experts recommend including representatives from each area when planning and deploying tools. This cross-functional approach ensures that different needs are addressed. It also helps create buy-in across the organisation, which increases adoption.

For example, at San Joaquin General Hospital, involving clinicians from all departments helped improve documentation and patient handoffs.

Address Equity and Bias Head-On

AI can improve care, but it can also reflect bias in data. Leaders must ensure AI tools work fairly for all patients. This means reviewing datasets and monitoring for any unfair behaviour from the tools.

A recent initiative by health advocates emphasises “equity-first” standards for healthcare AI to prevent bias and ensure fairness in care. Hospitals should adopt these principles to avoid deepening health disparities.

Communicate Clearly Across the Organisation

Change is hard. People resist what they don’t understand. Clear communication is essential.

Leaders should explain:

Why the organisation is adopting AI
What problems does it solve
How it helps clinicians and patients
What safeguards are in place

Communication builds trust. It reduces fear and increases willingness to use new tools.

Plan for Change Management

Adopting AI is not a one-time task. It’s a long journey. Leaders need a strong change management plan to guide teams through the transition.

Change management includes:

Training and support programs
Feedback loops for staff
Regular updates and improvements
Celebrating small wins

When teams feel supported, they adapt faster and work more collaboratively with AI.

Prepare for the Long Term

Healthcare leaders should think long-term, not short-term. AI adoption is not a quick fix. It’s a strategic shift that requires planning, patience, and persistence.

At Harvard, experts advise developing an AI framework that accounts for risks, compliance, and patient safety. This framework guides adoption from pilot to full implementation in a thoughtful way.

Planning for the long term also means investing in data systems, talent, and governance structures. Leaders don’t need to be technology experts—but they do need to know how to lead teams through this shift.

Use AI to Free Human Time for Care

One of the most significant benefits of AI is that it can take over repetitive tasks. This helps clinicians spend more time with patients.

For instance, some systems draft responses to patient messages based on medical records. Clinicians can review and send these messages quickly. This reduces after-hours work, sometimes called “pyjama time,” and brings time back to patient care.

This is what human-centred AI adoption looks like: AI giving clinicians time back so they can focus on what matters most—people.

Your Role as a Healthcare IT Leader

If you are bringing AI into your organisation:

Start with real problems, not cool tools.
Pilot early and learn fast.
Build trust with training and communication.
Create strong governance and ethical oversight.
Involve clinicians and staff in every step.
Measure outcomes with clear metrics.
Prepare for long-term adoption as a strategic shift.

These actions help ensure AI supports people first, not just technology.

Conclusion: Human-Centred AI Is the Future of Care

Human-centred AI adoption is not easy. But it is possible. With thoughtful strategy, strong leadership, teamwork, and clear goals, healthcare organisations can use AI to help patients and staff genuinely.

This approach turns AI from a buzzword into a fundamental tool for improving care and making life better for clinicians and patients alike. When leaders put people first, AI does more than automate—it enhances humanity in healthcare.