No One Trusts AI in Healthcare

2 min readDevraj Gopal

No one trusts AI.

I just returned from the 2025 American Hospital Association Leadership Summit in Nashville, and one theme rang louder than any keynote: trust is the new currency for AI in hospitals.

In healthcare, where mistakes are life-and-death, trust is more important than any other industry.

Meanwhile, the healthcare industry is finally hungry for AI. Meeting with top leaders, I saw the energy and optimism for what generative models can do across staffing, quality metrics, and more. Recent legislation has shown leaders that transformation is crucial for survival, with massive cuts looming on the horizon.

Despite the excitement, there’s still too much uncertainty. Nearly no one fully trusts AI solutions due to very real hallucination and liability concerns. Many clinicians push for AI out of necessity, not excitement.

The tension is real.

Over 80% of health‑system executives expect generative artificial intelligence to reshape their organizations this year. Most of them are still asking regulators for clearer guardrails.

Only 38% of clinicians trust AI systems to meet real-world needs, and 75% are concerned about the huge liability AI can bring.

How can we escape this distrust of AI?

I think it starts with building systems where trust is central, not just an add-on. This looks like transparent data pipelines, constant validation to prevent hallucination, and humans in the loop at every crucial decision. That’s just the beginning.

I’m excited for the future of trust and governance in healthcare AI.