Since the recent shooting of former UnitedHealthcare (UNH) CEO Brian Thompson in New York City, the health insurance provider has been under severe scrutiny. While the motives behind Thompson’s assassination remain unknown, the company is in full focus as the media shines an intense spotlight on its history and operations.
🚨Don’t miss this amazing Cyber Week Move! Get 60% off TheStreet Pro. Act now before it’s gone. 😲
Reports indicate that Thompson, a respected veteran of the health insurance field, had received threats prior to the attack. But as this is fairly common for leaders in his industry, Thompson and his company disregarded them. As of now, very little is known regarding the shooter’s identity or motives.
As details become available, though, reporters have been digging into UnitedHealthcare, the insurance arm of UnitedHealth, providing detailed context on an industry that doesn’t often receive much mainstream focus. One recent discovery highlights something that could be highly concerning for both investors and customers and could have wide-ranging implications.
Shutterstock
UnitedHealthcare’s major operational flaw
It’s no secret that artificial intelligence (AI) has taken over many industries. Since the launch of ChatGPT in 2022, companies have rushed to implement AI systems as a means of optimizing production and streamlining efficiency. But in the case of UnitedHealthcare, it seems to have caused some significant problems.
Related: UnitedHealth executive killed in 'targeted attack' in Manhattan
One month before Thompson’s assassination, hospitals began reporting an uptick in denied insurance claims. This prompted speculation that the AI bots being used by insurance providers were responsible. Families of now-deceased patients who UnitedHealthcare denied coverage opted to take legal action.
That’s hardly where it began, though.
A class action lawsuit filed on November 14, 2023, in the US District Court in Minnesota alleged that UnitedHealth Group and its subsidiaries, UnitedHealthcare and Navihealth, were guilty of using AI technology “in place of real medical professionals to wrongfully deny elderly patients care.”
According to the suit, the companies were aware that the AI model had a 90% error rate in evaluating their claims and had overridden the determinations made by both patients’ physicians.
The plaintiffs also accused UnitedHealthcare’s leaders of proceeding with its “flawed AI model” despite knowing its limitations because they also knew only a fraction of policyholders, roughly 0.2%, would appeal their claims, even after being denied the coverage they needed.
Lawsuits can take months to unfold, and so far, none of the claims made against the company have been proven. But if the accusations made by the plaintiffs are correct, it could indicate a pivotal moment for AI, specifically one of the new technology’s first public failures.
AI models and chatbots are consistently revolutionizing many industries, helping push us closer to self-driving cars and highly accurate weather predictions. However, these reports reveal that UnitedHealthcare’s model may have created significant problems for the many customers who rely on the health insurance provider’s coverage.
Related: Big Tech leader unveils the ‘ChatGPT of Weather Prediction’
In March 2024, a report from JAMA Network, a medical journal published by the American Medical Association, examined this topic, stating that “[the] use of machine learning methods exacerbates some problems with coverage denials, especially the opacity of decision-making.”
The authors noted that this could make it difficult for the patients whose claims have been denied to contest them afterward, as they likely won’t know the basis for the algorithm's verdict and cannot explain why it is flawed.
More Tech Stocks:
- Analysts reboot C3.ai stock price target on Microsoft deal
-
Tesla replaced by much smaller rival in EV charging project
- Nvidia may make a game-changing AI chip move
The implications of this public AI failure
The key takeaway from this story should be that no matter how impressive AI appears, the technology has limitations. On top of that, there may be some things it shouldn’t be trusted with, at least not yet. AI has proven helpful when it comes to robotically assisting surgeries. Still, when it comes to evaluating the claims that can determine the medical care someone receives, it is clear that the technology is deeply flawed.
Related: Apple unveils AI decision that is a major blow to Nvidia
With the health insurance industry under heavy scrutiny, investors will likely be looking to scale back their investments in the sector. Many companies have seen their stocks boosted by news that they are incorporating AI into their operations. However, the case of UnitedHealthcare is a good reminder that AI can pose a negative impact in some areas, at least until these models show significant improvement.
UnitedHealthcare isn’t the only company that uses AI for claim evaluation. Most of its competitors have done the same, likely seeing similar results. UnitedHealthcare rivals Cigna CI and Humana HUM have also been accused of denying medical claims through AI and have faced legal action.
AI hype is as strong as ever and isn’t likely to slow down, but it is clear that its effectiveness varies by sector.
Related: Veteran fund manager sees world of pain coming for stocks
