Entrepreneur Mark Cuban is warning that artificial intelligence could deepen some of the U.S. healthcare system’s administrative problems if insurers and large healthcare companies use competing automated agents to delay or deny care.
Cuban made the argument in a July 12 exchange on X after venture capitalist Marc Andreessen promoted the idea that AI systems are approaching or exceeding physicians on some medical evaluations. Cuban did not dismiss the technology’s clinical potential. Instead, he focused on the incentives surrounding payment, prior authorization and claims.
His concern is that doctors may gain powerful AI assistants while insurers and other intermediaries deploy their own systems to scrutinize requests, generate objections and extend appeals. In that scenario, automation could make an already complicated process faster without necessarily making it fairer or easier for patients.
Cuban also argued that employers often lack a clear view of the underlying price of care purchased through health plans. He has advocated for employers to contract more directly with medical providers and reduce reliance on intermediaries whose financial incentives may not align with patients.
Prior authorization already consumes time
Recent physician survey data show why the issue resonates. In a 2025 American Medical Association survey of 1,000 practicing physicians, respondents reported handling an average of 40 prior authorization requests per week. They estimated that physicians and their staff spent about 13 hours each week on that work.
The survey also found that 95% of respondents believed prior authorization delayed access to necessary care. Seventy-four percent said denials had increased over the previous five years, and 60% were concerned that insurers’ use of AI could increase denial rates. Those figures reflect physicians’ reported experiences and concerns; they are not an independent audit of every insurance decision.
Federal rules are beginning to impose tighter standards. The Centers for Medicare & Medicaid Services says affected payers must respond to urgent prior authorization requests within 72 hours and standard requests within seven calendar days beginning in 2026. New electronic data-exchange requirements are scheduled to take effect in 2027.
AI could also reduce paperwork, flag unsafe drug combinations and help clinicians locate relevant information more quickly. Cuban’s warning is therefore less about whether AI belongs in medicine than about who controls it, what goals it is given and whether patients can understand and challenge its decisions.
For patients, any automated recommendation still operates within a human healthcare and insurance system. Medical decisions should remain between patients and qualified clinicians, while coverage denials should include a reason and information about available appeal rights.
Sources: original report via Apple News, TheStreet, the American Medical Association and the Centers for Medicare & Medicaid Services.
