How AI Algorithms Decide What Health Insurance Will Cover—and Why Patients Are at Risk

How Artificial Intelligence Controls the Quality of Health Insurance Coverage in the U.S.

Over the past decade, health insurance companies have increasingly turned to artificial intelligence (AI) algorithms. Unlike AI tools used by doctors to diagnose or treat patients, insurers use AI to determine whether to approve payment for recommended medical treatments and services.

A common example is prior authorization, where a doctor must get approval from the insurance company before providing specific care. AI algorithms assess if the requested treatment is "medically necessary" and decide the amount of care a patient is eligible for, such as the length of hospital stays after surgery.

If an insurer denies coverage, patients usually face three choices:

• Appeal the decision, which often requires time, money, and expert help—only 1 in 500 denials are appealed.
• Accept an alternative treatment that the insurer will cover.
• Pay out-of-pocket for the recommended care, which is often unaffordable.

Insurance Algorithms: Balancing Efficiency and Patient Impact

Insurance companies claim AI helps them make faster, safer decisions about necessary care and prevents wasteful spending. But evidence suggests these algorithms can delay or deny care that should be covered, prioritizing cost savings over patient needs.

Insurers feed patient health records and other data into AI systems, which compare this information with medical standards to approve or deny claims. However, insurers refuse to disclose how these algorithms operate, making it impossible to verify their fairness or accuracy.

Automated reviews reduce the need for medical professionals in the approval process, saving insurers time and money. But denying valid claims quickly often leads to lengthy appeals, which can drag on for years. For seriously ill patients, delays may mean the insurance company avoids payment if the patient passes away before resolution.

A Pattern of Withholding Care

This approach raises concerns that AI is used to withhold coverage for expensive, long-term, or terminal conditions, including chronic illnesses and disabilities. Research shows patients with chronic diseases face higher denial rates and worse outcomes.

Disparities are evident as well: Black, Hispanic, and other nonwhite patients, along with LGBTQ+ individuals, are more likely to experience claim denials. Additionally, prior authorization may inadvertently increase overall healthcare costs by delaying necessary treatment.

Insurers argue patients can always pay for denied treatments themselves, but this ignores the financial reality for many. Denied or delayed care can have severe health repercussions for those who can't afford to pay.

Regulatory Gaps and Emerging Efforts

Unlike medical AI tools, insurance coverage algorithms face little regulation. They are not subject to Food and Drug Administration (FDA) review, and insurers often label their algorithms as trade secrets, keeping decision processes opaque.

There are no independent, peer-reviewed studies verifying these AI tools’ safety, fairness, or effectiveness in real-world insurance decisions.

The Centers for Medicare & Medicaid Services (CMS) recently announced that Medicare Advantage plans must base coverage decisions on individual patient needs rather than generic checklists. However, insurers still define their own standards for "medical necessity" and are not required to have their algorithms independently tested.

State-level efforts are emerging. States like Colorado, Georgia, Florida, Maine, Texas, and California have proposed or passed laws regulating insurance AI. For example, California’s 2024 law requires licensed physicians to supervise insurance coverage algorithms. But most laws still allow insurers significant leeway in defining standards and do not mandate third-party review before deploying AI systems.

State laws also cannot regulate Medicare or insurers operating beyond state borders, leaving gaps in oversight.

The Case for FDA Oversight

Many experts argue that the widening gap between insurers' cost-saving measures and patient health demands stronger regulation of coverage algorithms. The FDA is uniquely qualified to evaluate these tools because it already reviews many medical AI devices for safety and effectiveness.

FDA oversight could provide a consistent national regulatory framework, avoiding a patchwork of state rules that complicate compliance.

Currently, the FDA’s authority might be limited because it regulates medical devices intended for diagnosis or treatment, and insurance algorithms do not fit this definition. To empower the FDA, Congress may need to amend the law to include insurance AI tools.

Meanwhile, CMS and state governments could require independent, unbiased testing of insurance algorithms to ensure they are safe, accurate, and fair. This pressure might encourage insurers to support a unified national standard like FDA oversight instead of facing fragmented regulations.

Conclusion

AI is playing an increasingly critical role in health insurance coverage decisions, but current oversight is insufficient. Without transparency and regulation, patients risk denied or delayed care with serious health consequences.

Efforts to regulate insurance AI are underway but require stronger commitments to protect patients and ensure fair, evidence-based decisions. For professionals in healthcare and insurance, understanding these developments is essential as AI continues to shape coverage practices.

For those interested in learning more about how AI is transforming healthcare and insurance industries, exploring specialized training can provide valuable insights and skills. Check out Complete AI Training for up-to-date courses tailored to AI applications in healthcare and insurance.