Building trust in AI: Why health insurers must lead with transparency and human empathy

Artificial intelligence is quickly becoming embedded in nearly every part of the health care ecosystem, from helping clinicians sift through clinical documentation to supporting health plans as they streamline administrative processes and reduce the burden on members and providers. But for AI to improve healthcare at scale, the conversation must shift from what AI can do to how it can be trusted to do it responsibly.

Recent research from a recent study, across both clinicians and consumers, reveals a consistent theme: people embrace technology when they understand it, can question it, and see that human judgment stays firmly in the loop. Trust is not a nice-to-have. It is the prerequisite for adoption.

A new trust equation for health care AI

Today, payers and providers are using AI primarily to improve operational efficiency and accelerate decision-making. These benefits are real. But the next stage requires demonstrating how AI strengthens fairness, enhances clinical quality, and protects the human relationship at the center of care.

A key insight from a provider-focused study is that transparency alone is not enough, confidence calibration matters. When clinicians can see not only why an AI model made a recommendation but also how confident it is in that recommendation, their trust increases dramatically. In a study of 6,689 cardiac cases, a physician'soverride of the AI recommendation dropped from 87% to 33% once the AI’s confidence level was made explicit. When the model signaled high confidence, override rates fell to just 1.7%.

This reflects a simple truth: clinicians trust AI that behaves less like a black box and more like a colleague.

Human-centered AI in practice

For health plans, some of the most promising applications of AI are those that lighten the load on clinical reviewers and care managers. AI can:

• Scan hundreds of pages of medical history in seconds
• Surface the most relevant information
• Highlight documentation gaps early
• Reduce repetitive administrative work

These capabilities free clinicians to do what only clinicians can: apply judgment, interpret context, and make the final call.

In this model, AI does not replace the human; it equips the human.

Transparency as a driver of confidence

In addition to our provider research, we have also surveyed consumers. That research shows strong support for transparent, plain-language communication about when and how AI is used, for health plan functions such as prior authorization and claims processing. People want clarity, not complexity. They want to know that:

1. AI is an assistant, not a decision-maker
2. A clinician remains accountable for final determinations
3. They have easy access to explanations and appeals
4. Safeguards are in place to ensure fairness

This growing appetite for transparency aligns with new regulatory momentum at both the federal and state levels. Leading health plans are not waiting—they are voluntarily adopting clearer disclosure, stronger oversight, and more accessible member communication.

A blueprint for responsible adoption

As AI becomes woven into critical health care workflows, payers can lead by building trust through four commitments:

1. Clear, human-centered communication: Explain AI-enabled steps in plain language, with emphasis on the human role.

2. Accessible pathways for questions and appeals: Ensure members can easily ask for clarification or seek a second review from a clinician.

3. Ethical governance and oversight: Create multidisciplinary teams that oversee algorithms, audit performance, and evaluate clinical impact.

4. Continuous monitoring and improvement: Track model accuracy, bias, and reliability—and share progress with stakeholders.

Trust Is the foundation for AI’s next chapter

Across consumers, clinicians, and health plans, one message stands out: AI earns trust when it operates transparently, complements human expertise, and delivers tangible improvements in care quality and experience.

The future of health care AI will not be defined by the power of the algorithms, but by the confidence that patients, providers, and payers place in them. That confidence is built through openness, empathy, and a steadfast commitment to responsible use.

As the industry enters this next phase, the mandate is clear: lead with transparency, keep humans at the center, and ensure AI becomes a tool that strengthens—not replaces—the clinical judgment that healthcare relies on.

Sundar Subramanian is Chief Executive Officer at Zyter|TruCare. He previously spent over 16 years at PwC, where he led the strategy consulting business across industries (Strategy&) and the Enterprise Strategy, Value, and Digital Transformation practice, advising Fortune 500 companies on growth strategy, AI-enabled transformation, and operating model redesign. Prior to PwC, he led the Healthcare Services vertical and the Medicaid and Medicare Center of Excellence at Booz & Company and held senior roles at WellCare Health Plans, and McKinsey & Company.