By Editorial Team 
Legal challenges are piling up where artificial intelligence meets healthcare reimbursement, as payers roll out advanced systems to manage their claims queues. Right after the big lawsuit over UnitedHealth’s nH Predict tool, a new class-action case targeting Cigna’s PxDx system is getting a lot of attention from regulators, investors, and the people running health plans. That litigation is still active in the U.S. District Court, Eastern District of California, and it’s become a critical case study for anyone involved in AI-driven claims processing.
The Allegations Against Cigna’s PxDx System
The class-action lawsuit, Kisting-Leung v. Cigna Corp., was filed in July 2023 and it makes some pretty serious claims about Cigna’s PxDx system. The complaint’s central argument is that Cigna used PxDx to automatically deny a huge number of medical claims with almost no one actually looking at them. Plaintiffs argue this system basically automated medical necessity decisions that are supposed to be nuanced, letting Cigna deny payment requests in bulk. A May 2023 ProPublica investigation, which the lawsuit references heavily, is at the heart of these allegations. That report found that Cigna’s PxDx system was used to deny over 300,000 payment requests in just two months. Even more shocking, the ProPublica investigation into Cigna PxDx reported that Cigna’s own doctors spent an average of only 1.2 seconds reviewing each claim before denial. To top it off, the complaint points to one medical director who allegedly rejected something like 60,000 claims in one month, which naturally makes you question how legitimate those reviews could possibly be. These facts suggest a system built for speed-denials, not for careful medical review, which could mean patients were wrongly denied coverage for care they needed. For anyone who deals with regulatory compliance, the potential fallout is obvious: this could be a major breach of fiduciary duty.
Judge Drozd’s Procedural Ruling: A Glimpse into Judicial Scrutiny
We’re a long way from a final verdict, but a key procedural ruling came down on March 31, 2025. Judge Dale A. Drozd didn’t issue a final judgment, but he did allow some of the most important parts of the lawsuit to move forward. Specifically, Judge Drozd ruled that Cigna’s use of the PxDx system for medical-necessity decisions, without a medical director substantively reviewing each case, was an abuse of discretion according to the health plan’s own terms Judge Drozd’s ruling on Cigna PxDx. This part of the ruling is a big deal for health plan governance and compliance folks. It shows that courts expect medical necessity decisions to involve real human judgment, even when technology is involved. By letting the breach-of-fiduciary-duty claims proceed, Judge Drozd sent a clear signal that just handing off these decisions to a black box, without the human oversight promised in the plan documents, is legally risky. While other claims were dismissed (plaintiffs got 21 days to amend), the court is clearly taking a measured look. This procedural opinion stands in stark contrast to the “it’s just a guide” defense we’ve seen in other cases, like how UnitedHealth tried to frame its nH Predict tool. Judge Drozd’s focus on the plan’s requirement of a “substantive review” by a director could set a standard for how courts look at AI in claims.
Cigna’s Rebuttal: Not AI, Not an Algorithm
Cigna’s defense is pretty strange. Instead of admitting they use AI and arguing it’s just a helpful tool, Cigna’s official position is that PxDx “is not powered by AI or an algorithm” Cigna’s official statements on PxDx. This is a big departure from the normal conversation about AI in healthcare and it attacks the whole premise of the lawsuit. Cigna claims the PxDx system only looks at about 50 low-cost procedures and tests, and that it only does so after the patient has already gotten the treatment. They say PxDx is just a standard review process, a lot like the ones the Centers for Medicare & Medicaid Services (CMS) and other payers have been using for years. Their defense paints PxDx not as a smart AI, but as a more old-fashioned, rules-based system for post-service review. Outlets like Healthcare Dive, Healthcare Finance News, InsuranceNewsNet, Bloomberg Law, NFP, and Employment Law Weekly have all reported on Cigna sticking to this story. For investors and compliance professionals, this distinction is everything. If PxDx really isn’t an AI or a predictive algorithm, then the legal fight shifts away from things like algorithmic bias and opacity. But the plaintiffs’ claims, backed by the ProPublica numbers, describe a system that denies claims so fast and in such high volume that you have to wonder about the level of human review, regardless of what you call the technology.
Implications for Responsible AI in Healthcare
The Kisting-Leung v. Cigna Corp. case, along with the nH Predict fight, is a sign of what’s to come for automated systems in healthcare administration. The conflict is between the efficiency that technology promises and the absolute need for patient safety, ethical choices, and following the rules. For governance and compliance teams inside health plans, the Cigna case is a wake-up call. You have to document exactly how your automated systems, whether you call them “AI” or just “rules-based”, fit into your workflows. Human oversight can’t just be for show. It has to be real, especially when a medical necessity decision is on the line. Judge Drozd’s ruling confirms that the language in your plan documents about a medical director’s judgment isn’t just fluff, it’s a real requirement that technology can’t bypass. Health tech investors, especially those looking at companies in the claims processing space, need to think about the legal risk. The Cigna case shows that even if your system isn’t technically “AI,” you can still get into deep legal trouble if it looks like you’re letting a machine make critical medical calls without proper human sign-off. The market is going to start preferring solutions that have clear audit trails and provable human-in-the-loop checks, not just ones that are fast. This lawsuit, with Cigna’s odd defense about what PxDx actually is, is going to force the courts and the industry to get much clearer about what “AI” means in a legal sense and what the rules are for automation in healthcare claims. It’s a reminder that you can’t just innovate your way around your ethical and legal duties.
Frequently Asked Questions
What are the core allegations against Cigna’s PxDx system?
The lawsuit alleges Cigna’s PxDx system systematically denied a vast number of medical claims with minimal human oversight. Plaintiffs contend the system automated medical necessity determinations, leading to rapid denials of payment requests.
What was the significance of Judge Drozd’s procedural ruling on Cigna’s PxDx system?
Judge Drozd ruled that Cigna’s use of PxDx for medical-necessity determinations without a medical director substantively reviewing each case constituted an abuse of discretion under the health plan’s terms. This allowed critical aspects of the lawsuit, including breach-of-fiduciary-duty claims, to proceed, emphasizing the judiciary’s expectation of substantive human judgment in such decisions.
How does Cigna defend its PxDx system against the allegations?
Cigna claims PxDx is not powered by AI or an algorithm and reviews only about 50 low-cost tests and procedures after treatment has been received. They assert PxDx functions as a traditional, rule-based system for post-service review, similar to processes used by CMS and other payers for years.
What is the potential impact of this lawsuit for health plan governance and compliance teams?
The ruling underscores that medical necessity determinations, even with technological tools, must be rooted in substantive human judgment. It signals that delegating critical decisions solely to an algorithm without adequate human oversight, as defined by plan terms, could be legally problematic and a breach of fiduciary duty.
