Blue Shield CEO urges AI and digital records to replace insurer mandates

Blue Shield CEO urges AI and digital records to replace insurer mandates

Paul Markovich, president and CEO of Ascendiun—the nonprofit parent company of Blue Shield of California, Blue Shield Promise Health Plan, Altais, and Stellarus—is calling for a fundamental shift in how medical treatment plans are determined. Markovich argues that healthcare decisions should be moved away from insurance bureaucracy and returned to the patient and physician through a "shared decision-making" approach.

Central to this vision is the elimination of prior authorization, the process where doctors must wait for an insurer to determine if a treatment plan is covered. Markovich describes this exercise as "dreaded" and envisions a system where patients experience a hassle-free process from diagnosis to recovery without spending hours or days waiting for insurance approvals.

The Three Pillars of Reform

According to Markovich, making shared decision-making the norm requires three specific components:

• A comprehensive, real-time, digital health record accessible to both patients and clinicians.
• An objective AI model based on credible healthcare data to assist in diagnosis and treatment recommendations.
• Sufficient time for patients to have in-depth conversations with doctors who are free from financial conflicts.

Markovich notes that the shared decision-making model was introduced more than 30 years ago by John E. "Jack" Wennberg and colleagues. Their research indicated that when patients understand their options and discuss them with doctors, they tend to choose more conservative, less costly treatments while achieving better results and higher satisfaction.

AI Potential and Clinical Reality

The push for AI integration is driven partly by the impossibility of human doctors staying current with all research. Markovich cites a 2004 study calculating that epidemiology-trained doctors would need 29 hours per weekday to remain fully up to date on primary care research.

Technological leaps in 2026 suggest AI can fill this gap. Markovich points to Microsoft's AI Diagnostic Orchestrator, which correctly diagnosed 85% of medical cases in the New England Journal of Medicine, compared to about 20% for human doctors.

The practical impact of these tools is illustrated by the case of Aidan Brown, a cancer survivor diagnosed with stage 4 neuroblastoma. His mother, Michele Brown, previously described a grueling journey involving a binder of hundreds of medical records and spreadsheets to track symptoms as they moved between hospitals. Markovich suggests that if such treatment happened today, AI could analyze digital records, images, and lab results in minutes to identify key patterns and potential treatments, including clinical trials.

The Ethical Divide: Patients vs. Insurers

While Markovich sees AI as a tool for clinical empowerment, other experts warn that AI is already being used by insurers to restrict care. Isaac Kohane, a Nelson professor of bioinformatics at Harvard Medical School, highlighted an ethical dilemma where an AI model recommended human growth hormone for a patient when prompted as a pediatric endocrinologist, but provided a rationale for denying care when prompted as a health insurance representative.

This reflects a broader trend in the industry. A 2024 survey of 93 large health insurers across 16 states found that 84% were using AI for operational purposes. Some insurers have integrated these tools into utilization reviews to verify coverage requests. United Healthcare, for instance, now has more than 1,000 AI applications, though it faced a 2023 class action lawsuit accusing it of using a flawed AI algorithm to deny claims.

Research from the Stanford Healthcare Ethical Assessment Lab for AI suggests that while AI could automate the approval of the 93% of Medicare Advantage prior authorization requests that were approved between 2019 and 2023, there are significant risks. Professor Michelle Mello and colleagues write in Health Affairs that a lack of meaningful human review may lead to wrongful denials.

Systemic Barriers and Future Regulation

Markovich admits that AI and digital records are not enough alone. He argues for a broader rethinking of how providers are paid, how prescription drugs are priced and distributed, and how to prevent system abuse.