E-News Exclusive

AI and the Patient Record

By Selena Chavis

Industry professionals weigh in on AI’s growing influence in how patients access their health information.

For an industry that is typically reticent to adopt new technology or new care delivery models, health care is the surprising leader when it comes to AI deployment. Industry adoption of AI is experiencing a 36.8% compound annual growth rate and accounts for nearly half of all spend across other verticals listed in the top 10—approximately $1.5 billion in 2025.

For health care stakeholders, applications of AI are showing up everywhere to improve performance, from speeding revenue cycles and addressing routine administrative tasks to eliminating backlogs that slow diagnostics. The far-reaching influence of AI is also driving dramatic shifts in how patients mediate their own health information via such tools as ambient voice transcription, after hours use of chatbots, and AI-generated health summaries.

A March 2026 KFF Tracking Poll found that about one-third of US adults have turned to AI chatbots for health information in the past year alone. Among those users, 41% said they have uploaded personal medical information like test results or doctor's notes into an AI tool to get personalized explanations or advice. In addition, a 2026 physician survey from the American Medical Association (AMA) found that 19% of physicians are now using AI to generate draft responses to patient portal messages, alongside chart summary generation at 28% and discharge instruction drafting at 30%.

While the efficiency gains are real, industry professionals warn that the rapid shift in how patients access information is concerning. For some, the question of who is responsible when something goes wrong remains largely unanswered.

AI's Evolving Role in Patient Insights
The role AI plays in a patient’s health information is changing rapidly. What began as a tool for organizing and surfacing insights is evolving into something more consequential—an active interpreter of clinical data.

"Instead of just reviewing results, patients can now ask questions and receive simplified explanations in real time," says Stacey Rosenberg, DNP, RN, ACNS-BC, CNE, associate chief nursing administrator at South New Hampshire University (SNHU). “I experienced this when an AI tool interpreted my own mammogram report. It translated clinical language into something more understandable, but it also highlighted how much context can be lost in translation. That's the shift we're navigating—from access to interpretation, often without the full clinical picture."

For patients navigating an already complex system, real-time AI interpretation can be genuinely valuable. For example, a 2026 Wolters Kluwer Future Ready Healthcare report found that 70% of patients and clinicians agree that AI is enabling better patient health literacy and engagement, with 52% of patients using AI to research health conditions before clinical encounters.

At the same time, those same features that create fast, accessible information can strip out the clinical nuance that makes a record accurate and actionable. A 2025 PLOS Digital Health study at the University of California, San Francisco, demonstrated the impact by having AI models summarize 100 real emergency department encounters. Findings suggest that while the models produced clinically useful summaries, they also hallucinated and omitted details, particularly in the “plan” section where clinical accuracy is most critical.

Rosenberg notes that industry professionals need to be wary of the convenience trap. "The upside is clear. AI makes health information more accessible and easier to engage with. The challenge is that simplification can quickly become distortion,” she points out, adding that even when summaries are accurate, they can leave out important context that might make uncertain findings sound more definitive or create a sense of certainty that is not warranted.

The organizations navigating this best, according to Lyndsay Goss, DNP, RN, CNE, NPD-BC, director of continuing professional development at SNHU, are the ones keeping humans in the loop. "The organizations getting this right are using AI to support, not replace, clinical communication," she emphasizes.

The Accountability Question
As AI becomes more embedded in how patients receive and interpret their health information, the legal and ethical stakes naturally rise. According to Rosenberg, the question of accountability remains complex.

“If a patient receives inaccurate AI-generated guidance and acts on it, potential liability could involve the health care organization, the clinician overseeing care or the technology vendor,” she says.

Right now, guidance from organizations like the AMA highlights liability, transparency, and oversight as key concerns, especially as AI becomes more embedded in care delivery. Ethically, the issues go deeper, Goss notes, adding that patient safety remains the top priority, and inaccurate AI-generated information has the potential to directly influence clinical decisions in ways that could be harmful.

"Health systems are still responsible for the tools they deploy, and clinicians remain responsible for how information is interpreted and communicated," Goss says. "AI can support understanding, but it cannot replace clinical judgment, especially when patients are making decisions based on what they believe the information means."

Texas, California, and Colorado have all enacted disclosure requirements mandating that patients be informed when AI influences their care. Rosenberg sees this kind of transparency as nonnegotiable.

"Patients and clinicians deserve to know how information was created, especially if AI played a role. Health systems shouldn't hide AI use in the background," she says. "A patient portal note could include a statement like, 'This summary was generated using AI and reviewed by your care team.' We don't want to overwhelm people with technical details, but we do want to build trust. Trust comes from being honest, clear, and consistent."

HIM professionals are an important part of the accountability question going forward, Goss says. As AI generates, summarizes, and interprets more of the clinical record, the HIM profession is being called to evolve from records custodian to AI governance steward.

"HIM professionals are going to be central to this entire conversation," Goss says. "AI in health care is only as good as the data behind it. And that's where HIM expertise comes in: data governance and integrity; privacy, security, and compliance; quality assurance and auditing; AI policy development; and education and leadership."

— Selena Chavis is senior director of accounts with Insenna and a Florida-based freelance writer.