Late last month, the Trump administration announced that it had cut 760,000 people from the Affordable Care Act rolls, claiming fraud and improper enrollments and touting billions in potential taxpayer savings.
Check out the NPR Politics Podcast for daily analysis of the biggest news out of Washington, D.C.
Not all of those kicked off were fraudulently enrolled, KFF Health News has learned. Insurance industry sources said the Centers for Medicare & Medicaid Services has reinstated a small number of people whose coverage should not have been canceled. More could follow in the coming weeks if enrollees discover they are uninsured when they go to use their coverage.
A fix for widespread fraud?
Still, the sweep was long overdue, say supporters such as Brian Blase, whose Paragon Health Institute has supplied many of the sometimes controversial findings cited by those who allege widespread enrollment fraud in the ACA. But other policy experts question how many legitimate consumers were caught in the dragnet, perhaps because they failed to open a letter from their insurer or answer a phone call.
Questions continue to swirl, even as a second phase of the anti-fraud effort is underway. An additional 419,000 sign-ups are being scrutinized. Insurers must let CMS know how many of those enrollees don't respond to outreach efforts by "around mid-to-late November, after the midterm elections," said David Merritt, senior vice president of external affairs at the Blue Cross Blue Shield Association.
He confirmed that some BCBS members are among those who have been reinstated. He also said that CMS has an expedited review process for those who discover they have been improperly disenrolled.
Few details
Overall, however, the administration is providing few details about the cancellations, such as how robust the outreach was to those it targeted, how long it takes for people to get reinstated after discovering their coverage was wrongly canceled, and how many such cancellations have been reversed.
"It may be true that some of those enrollments were not authorized, but it can be equally true that people got lost in the shuffle and have a valid enrollment. Without more information, it's almost impossible to tell," said Katie Keith, a director of the Center for Health Policy and the Law at Georgetown Law. "That they are not being very transparent is telling too," she said of the administration.
The enrollment cut comes as part of a larger Republican strategy to focus on fraud in government programs, including Medicaid and hospice care, which polls show resonates with GOP voters.
At an Oct. 14 press conference, Health and Human Services Secretary Robert F. Kennedy Jr. and CMS Administrator Mehmet Oz touted the removal this year of subsidies for about 189,000 ACA enrollees with duplicate Medicare, Medicaid, or Children's Health Insurance Program coverage, citing not fraud but a failure to look for concurrent enrollments or continued coverage for dead people. They credited an effort begun last year by the Trump administration to restore more frequent data matching checks than were done under the prior administration.
Unauthorized sign-ups by rogue insurance agents seeking commissions and other potential impropriety in the ACA marketplace — applicants' providing incorrect income estimates, for example — has been a concern for several years, with much debate over the magnitude of the problem.
Even so, how the move will play out politically is uncertain. Healthcare costs are tied with gas prices as a top concern for voters. It is likely that many removed from the ACA are in red states, such as Texas, Florida, Louisiana, and Mississippi. Those areas saw rapid growth in ACA enrollment from 2020 to 2025, driven in part by now-expired enhanced subsidies put in place during Joe Biden's presidency. Additionally, the cancellations target only states that use Healthcare.gov, the federal ACA marketplace. The 21 states fully running their own marketplaces generally have more robust fraud prevention programs and have reported few or none of the kinds of unauthorized sign-ups plaguing the federal marketplace.
"To the extent there is a reaction depends on whether the government got it wrong, whether people try to get care and then find out their coverage is no longer active," said Adrianna McIntyre, assistant professor of health policy and politics at the Harvard T.H. Chan School of Public Health.
How it played out
According to information the agency provided to KFF Health News, CMS in June selected for further review nearly 1.2 million enrollments that had all of what it deemed four warning signs for fraud:
- The application included no Social Security number or immigration document number.
- A broker was involved in the sign-up process.
- The plan had no monthly premium payment.
- The enrollee had had no contact with their insurer and no claims filed.
Then, as first reported by KFF Health News, CMS sent lists of those ACA accounts to specific insurers, instructing the carriers to review them and use at least two methods of contact — such as a phone call and a letter — to try to reach the enrollees. Policyholders were given 30 days to respond to the insurers' outreach and provide the missing information. CMS disenrolled those who didn't respond.
Johnny Wallin, chief operating officer of Texas-based Marketplace Insurance Exchange, which helps clients in about 35 states enroll in coverage, said insurers reached out to his organization for help contacting people on the CMS lists, prompting "an all-hands-on-deck" effort by agents to reach clients, some of whom were vacationing overseas.
"Getting them to answer the phone isn't easy," he noted, although he suspects his agents had a better chance of getting a reply than an insurer because their clients are familiar with their phone numbers. Still, they could not reach about 60 of them.
Additionally, his team has been asked by insurers to help contact those targeted in the second phase of CMS' review. A recent case involved a client who needed to upload his new green card after his previous one expired, but language barriers meant that he had to take time off work to come in for agents to help with the process.
Information from a financial filing
CMS would not provide a list of the insurers it contacted, but publicly available documents show that New York-based Oscar Health was one.
In an August earnings call with analysts and investors, CEO Mark Bertolini said the company was reviewing the files and that "there are a number of cases where we've actually had contact" with people on the list. In other words, the company reached enrollees CMS had targeted over fraud concerns.
He did not specify how many it had reached. The company declined to comment for this article because it is in a voluntary "quiet period," when publicly traded companies limit disclosing information before the release of an earnings report.
Still, in a filing with the Securities and Exchange Commission, Oscar Health estimated that 250,000 to 300,000 of its members might be disenrolled because of the CMS effort. Nationally, the company had about 3 million members in the second quarter and reported strong profits.
Overall ACA enrollment has declined sharply, falling by nearly 3 million in the first few months of the year, due in large part to the expiration at the end of last year of the enhanced subsidies. The Trump administration has also taken some credit, saying its focus on weeding out improper enrollments and other changes, such as eliminating a special year-round low-income enrollment period, led to the decline.
CMS has also suspended some brokers for improper activity and barred new brokers from enrolling clients in ACA insurance on the federal exchange during the upcoming open enrollment period.
CMS did not respond to requests for data on the reasons it canceled enrollments. What percentage of enrollees were deemed ineligible due to incorrect income reporting or because they had other health coverage, such as through their jobs? How many because their ACA applications lacked immigration documents or Social Security numbers?
"Currently, there is no additional data to share," the agency's press office said.
But people outside the administration have been vocal.
"This was clearly phantom enrollees," said Paragon's Blase. He thinks more than 99% of the cancellations are of zero-premium coverage that people were signed up for by rogue call centers without their knowledge, or for people who don't exist at all.
Blase said the carriers had a big incentive to get hold of the enrollees because they would lose the subsidies they were being paid if the plans were canceled.
And for any people unfairly caught up, Blase and Trump administration officials say all they have to do to be reinstated is call the marketplace and provide their Social Security number or other missing information.
"Everyone has the political incentive for that process to be that easy," Blase said. "We're a month before the election."
But, as prior reporting has shown, it isn't always a quick process to get an ACA plan canceled or reinstated following an unauthorized sign-up or switch, even when a consumer discovers it themselves and reaches out to the marketplace.
Brokers and policy analysts note that some of the factors used to question enrollments — such as a plan's having no claims filed in a year — are not that unusual. Many people, for example, don't visit a doctor or hospital in a given year.
Insurers "want those healthy people in the insurance pool to help bring down the rates," said Ronnell Nolan, president and CEO of Health Agents for America, a trade group for health insurance agents. "We're against fraud, and CMS has allowed a lot of fraud to happen. But we don't want people who just didn't go to the doctor to be canceled."
And even with the outreach required by CMS, not everyone answers phone calls from unknown numbers or opens their mail or emails.
A recent marketing report found an average "open rate" on emails of 28.6%.
Policy experts also note that, while lack of a Social Security number on an ACA application could indicate a fake enrollee, it might also be an oversight, or the number might not be required, like for a child if the person applying is an adult who is not seeking coverage for the child.
"There will be legitimate people who don't have a claim or don't respond to an inquiry from an insurer," said Matthew Fiedler, a senior fellow at the Brookings Institution. "If the goal is to target fraudulent enrollment, what CMS did is better than tossing people at random, but is likely to miss the mark."
KFF Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF — the independent source for health policy research, polling, and journalism.
Copyright 2026 KFF Health News