This voice is automatically generated. Please let us know if you have any feedback.
LAS VEGAS — CMS wants to deploy artificial intelligence tools to help Medicare beneficiaries better navigate their care, CMS officials said Thursday at the HIMSS conference.
The agency is already using this technology to detect fraud. But CMS also wants to get this technology into the hands of patients, both to help seniors and to keep patients healthy. It is expected to curb the rise in medical costs, which continue to outpace other economies.
“The fundamental problem now is that other sectors of the U.S. economy have advanced through the use of technology and are deflationary,” CMS Administrator Dr. Mehmet Oz said during a panel discussion. “Inflation in medical care continues.”
Oz said the federal government wants to deploy AI agents, or AI systems that can act more autonomously, to help Medicare beneficiaries find doctors and choose Medicare Advantage plans. That tool should be available “by the end of this administration,” he said.
The biggest challenge, Oz said, is that Medicare beneficiaries don’t trust AI.
A survey released last year by health policy researcher KFF found that only 31% of Medicare enrollees over the age of 65 have “a lot” or “a lot” of trust in AI to access their medical records and provide personalized health advice.
“No one has told them the use case of why it will change their lives for the better,” Oz says. “This looks like a tool that we use to market to hospitals and help hospitals address problems, but it doesn’t necessarily address hospital problems.”
One of the goals of CMS’ Health Tech Ecosystem, an initiative launched last year to facilitate data exchange and increase the availability of digital health tools through partnerships with the private sector, is to expand access to conversational AI to help patients navigate their care and discuss their health, Amy Gleason, acting administrator of the U.S. DOGE Service and senior advisor to CMS, said during a panel discussion.
He noted that some companies that have signed the health tech ecosystem pledge have begun releasing dedicated AI chatbots specifically for health-related questions. For example, OpenAI announced a tool in January, and Microsoft announced its Copilot Health assistant this week.
Still, experts say misuse of AI chatbots could pose a risk to patients, given that they can provide false or misleading information. A study of OpenAI’s consumer health tools released last month found that chatbots often poorly triage cases, failing to send patients to the emergency room for serious health issues such as impending respiratory failure.
Hallucinations are a concern, but doctors can also make mistakes, Gleason said.
“The same thing will happen with AI. AI is the stupidest thing ever today,” she says. “So I think we need to embrace this and help people understand how helpful this is.”
The Trump administration is also using AI in health care to help CMS detect fraud, said Kimberly Brandt, the agency’s chief operating officer and deputy administrator.
Fraud has become a key priority for the Trump administration, which last month withheld $259 million in Medicaid funds from Minnesota due to unsubstantiated and potentially fraudulent claims. The agency is also suspending Medicare coverage for some suppliers of durable medical equipment for six months as part of its efforts to crack down on fraud.
Brandt said CMS uses algorithms trained on past Medicare providers against whom the agency has taken enforcement action, and tests new providers enrolling in the program against bad actors.
“We wanted to leverage AI to figure out where the problem was,” she said.
