UnitedHealth Group Inc. is tracking how often some employees use artificial intelligence tools as part of a push to embed the technology throughout its operations, according to people familiar with the matter.
The company is monitoring whether some workers in its Optum services division perform at least one query a day with programs such as OpenAI's ChatGPT or Microsoft Corp.'s Copilot, said the people, who asked not to be named discussing internal operations.
UnitedHealth is investing in AI "to fundamentally transform how we organize, operate, and work as a company moving forward," a spokesperson said. The company has implemented more than one thousand AI use cases and boosted training on the technology for employees and external customers, the spokesperson said.
A company document on AI training reviewed by Bloomberg described an engagement dashboard to track usage, training progress and "identify adoption gaps." It also said the company's AI applications had avoided the need for more than 15 million calls, adjudicated hundreds of millions of claims, and contributed more than 150 million lines of code. The document also included positive feedback from people who had been through the training, called AI Dojo.
AI is central to UnitedHealth's plan to rebuild confidence on Wall Street after its profits collapsed last year. The healthcare conglomerate has touted AI's potential to improve the fragmented U.S. healthcare system. Last year, the company started testing a system called Optum Real that's intended to speed up how medical claims are processed.
The technology is "advancing even faster than people can appreciate or understand," Chief Financial Officer Wayne DeVeydt said at a recent conference. The company is focused on "speed and agility," he said. "We think there's a window here where we can lead the industry in ways that others cannot."
UnitedHealth is investing $1.5 billion in AI this year and is seeing at least a 2-to-1 return on those investments, with positive returns within the first year, DeVeydt said in April.
Analysts at Raymond James pointed to the potential for AI initiatives to increase the company's profits when they upgraded the stock to a buy-equivalent rating in April.
UnitedHealth's latest annual report included a new disclosure related to AI: "Our increasing use of AI presents legal, regulatory and business risks," the company said. It also cited the potential for "inaccurate, incomplete or biased output" to affect operations or financial results.
UnitedHealth has seen big technology strategies backfire before. When the company announced a $13 billion deal with Change Healthcare in 2021, it pledged the deal would "simplify core clinical, administrative and payment processes" to improve healthcare while cutting costs.
In 2024, Change was struck by a massive cyberattack that paralyzed healthcare payments. The hack, traced to an account that didn't have basic cybersecurity precautions turned on, ultimately exposed data from 190 million Americans, more than half the nation's population.