Dr. Mehmet Oz told reporters that CMS believes roughly 35 percent of Affordable Care Act enrollees “may not be legit,” a figure that could mean five to six million people are on the rolls without using coverage, and he tied the surge in enrollment to an agency decision to stop strict verification.

Centers for Medicare & Medicaid Services administrator Dr. Mehmet Oz laid out a sharp allegation at a White House briefing: a huge slice of ACA enrollment may be illegitimate. He put the estimate at roughly 35 percent, translating to an estimated five to six million people whose premiums taxpayers may be covering without any claims being filed. That claim has big implications for both program integrity and the federal budget.

Oz called this the most specific fraud allegation yet from a sitting CMS chief, and he wanted attention on the numbers rather than slogans. Enrollment on the exchanges was about nine million for years and now tops 20 million, he said. His argument is straightforward: the population didn’t grow enough to explain the jump, the verification process did not.

“Let’s take a step back; 2015, ’16, ’17, ’18, 2020, there were nine million people on Obamacare, nine million. Today, it’s more than 20 million… What happened was we completely took the guardrails off, and I know this because I’m working in the agency that actually was told to take the guardrails off.”

That phrase, “told to take the guardrails off,” frames this as a policy choice rather than an administrative mistake. Oz said the agency was instructed to relax checks, which then produced headline-friendly totals but weaker integrity. From a Republican point of view, that looks like trading oversight for optics and taxpayers pay the bill.

“Because there wasn’t an earnest desire to keep track of whether you were appropriately on it or not, but more importantly, just get the number to where we all feel good about ourselves, we had massive increases of people joining the program.”

The mechanism Oz described is easy to grasp: CMS found many enrollees who never used the coverage and never filed claims. Those records point to dishonest broker sign-ups and duplicate enrollments across states rather than newly insured people finally getting care. If true, this is not accidental waste; it is an exploitable gap in program design and enforcement.

“We believe that 35 percent roughly, of the people that are using the Affordable Care Act, Obamacare exchanges, because they never used the program once, they’ve never filed a claim, may not be legit. And that actual number may translate to… five, six million people we could be paying premiums for because they don’t have to contribute anything, so they don’t even know they’re getting it.”

Oz spelled out two recurring fraud patterns: Medicaid enrollees being signed up for ACA plans by brokers to chase commissions, and people appearing enrolled in full-subsidy plans in more than one state at once. Both patterns point to incentives misaligned with program integrity, where brokers benefit from sign-ups regardless of eligibility. That kind of perverse incentive demands both data audits and enforcement.

“These are people who have Medicaid, and someone often a broker dishonestly enrolling them in Affordable Care Act, or they’re in two states at once getting full insurance paid for by us in multiple states at once, so we have evaluated these numbers. They’re extremely concerning.”

The scale matters because tens of millions of dollars in premiums roll through the exchanges every month. If five to six million enrollees are phantom or duplicative, the fiscal damage compounds quickly. For Republicans focused on taxpayer stewardship, the possibility that millions were counted but never used benefits is a major policy failure to fix.

“We had tried to do this when we first came into office; a court enjoined us, did not feel we followed appropriate… APA guidelines, government guidelines, and so we redid it, and now went back out again.”

Oz said the administration attempted a rule to tighten verification soon after taking office but was blocked on procedural grounds, then revised and reissued the rule roughly two weeks before the briefing. The back-and-forth — rulemaking, judicial pause, rewrite — is familiar turf for regulatory change in recent years. The next test will be whether the revised rule survives legal challenge and actually cleans the rolls.

Vice President J.D. Vance’s Task Force to Eliminate Fraud has made CMS a focal point, and that political spotlight is no accident. Enrollment data sits where health care spending meets program integrity, and Republicans see this as proof that audits and enforcement matter. The political heat comes with a real oversight responsibility.

Important gaps remain in what Oz released: he did not lay out detailed methodology for the 35 percent figure, and absence of claims is not an automatic sign of fraud because some enrollees may simply be healthy. The administration will need to publish verified data and a clear methodology to back a claim of this magnitude. Until then, the number is a charge that demands proof.

Karoline Leavitt was on maternity leave during the briefing, so Oz became the face of the announcement and the administration’s public explanation. His medical background gives the claim visibility, but visibility is different from documentation. The agency must produce the receipts — public methodology, enforcement actions, and measurable cleanup — if this is to move beyond a political talking point.

There are more than 20 million people on the ACA rolls, a jump of roughly 11 million in recent years, and CMS now says five to six million of those enrollees may never have used the coverage they were signed up for. That set of facts frames a direct accountability question for federal health policy and for anyone who pays the bill.