The Growing Role of AI in Health Insurance
Over the past decade, health insurance companies have increasingly adopted artificial intelligence (AI) algorithms to determine whether to approve or deny coverage for medical treatments and services. While AI aids doctors and hospitals in diagnosing and treating patients, insurers use it to make critical decisions about payment and coverage.
One of the most common uses of AI in insurance is for prior authorization. This process requires doctors to obtain approval from an insurer before proceeding with a treatment. AI algorithms evaluate whether a proposed treatment is “medically necessary” and, therefore, eligible for coverage. These systems also determine the extent of care a patient should receive, such as how many days they can stay in the hospital after a procedure.
Limited Options for Patients
When insurers deny coverage based on AI assessments, patients are left with few alternatives. They can choose to:
- Appeal the decision, a time-consuming and costly process that few undertake—only 1 in 500 denials are appealed.
- Accept a different treatment that is covered by their policy.
- Pay out of pocket for the recommended treatment, which is often financially unfeasible.
These limited options can have serious repercussions, especially when medically necessary care is denied or delayed. As a legal scholar focused on health law and policy, I find the implications of AI-driven denial of care particularly troubling.
Concerns Over Transparency and Fairness
Insurers claim that AI allows for quicker, safer decisions that reduce wasteful treatments. However, there’s mounting evidence that these tools are sometimes used to delay or deny valid care to save money. Critics argue that this practice leads to unnecessary suffering, especially for individuals with chronic or terminal conditions.
Although AI systems presumably analyze patient records in conjunction with medical guidelines, insurers have not disclosed exactly how these algorithms function. This lack of transparency raises ethical and legal questions, especially since insurance companies are not required to reveal their decision-making processes, citing trade secrets.
Moreover, AI systems can be manipulated to create financial advantages for insurers. For example, if an algorithm quickly denies a claim and the appeal process stretches over months or years, the insurer may save money, particularly if the patient’s condition worsens or becomes terminal during that time.
Disparities in Coverage Denials
Studies show that vulnerable populations are disproportionately affected by AI-driven coverage denials. Individuals with chronic illnesses, as well as Black, Hispanic, and LGBTQ+ communities, face higher rates of claim rejections. These disparities highlight the potential for AI to exacerbate existing inequities in healthcare access.
Some evidence also suggests that prior authorization systems may increase healthcare costs rather than reduce them, contradicting one of their primary justifications.
The Push for Regulation
Unlike medical AI tools, which are subject to oversight by the Food and Drug Administration (FDA), insurance algorithms remain largely unregulated. They don’t undergo FDA review, and insurers are not obligated to submit these tools for independent testing or validation. This regulatory gap means that there is no assurance of their safety or effectiveness.
Recent developments suggest that change may be on the horizon. The Centers for Medicare & Medicaid Services (CMS) now require that Medicare Advantage insurers base decisions on individual patient needs. However, these rules still allow insurers significant leeway in defining standards and do not mandate external testing of AI tools.
Some states have begun to take legislative action. For instance, California passed a 2024 law requiring a licensed physician to oversee insurance algorithms. Other states like Colorado, Georgia, Florida, Maine and Texas have introduced similar bills. However, these efforts often fall short, as they continue to give insurers too much control and lack robust enforcement mechanisms.
The Role of the FDA and Congress
Many legal experts believe that the FDA is well-suited to regulate insurance algorithms. The agency already has the expertise to evaluate medical AI tools, and expanding its authority could provide a uniform national regulatory framework. However, current FDA definitions may not cover insurance algorithms since they’re not used to diagnose or treat disease.
For effective oversight, Congress may need to amend the definition of a medical device to include insurance algorithms. In the meantime, CMS and individual states could mandate independent testing for safety and fairness, pushing insurers toward supporting a cohesive national standard.
Until comprehensive regulation is enacted, patients remain at risk of being denied critical healthcare due to opaque and potentially biased algorithmic decisions. The move to regulate AI in health insurance has begun, but achieving meaningful oversight will require coordinated efforts at both the federal and state levels.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.