Nationwide Poll Shows Democrat Voters Turning on Their Party Over AI Fraud Hunting

Government fraud has drained hundreds of billions of taxpayer dollars while Washington did nothing to stop it.

A new nationwide poll just tested whether Americans still want that fraud stopped using artificial intelligence.

The results show a shocking share of Democrats breaking from their party's script on AI fraud hunting.

Poll Shows Republicans and Democrats Actually Agree on Something

Noble Predictive Insights surveyed 2,533 registered voters between August 12 and August 16.

The sample split almost evenly, 930 Republicans and 930 Democrats, with 673 independents rounding it out, and a margin of error of just 2 percentage points.

Sixty percent said the government should use AI to hunt down fraud, waste, and abuse in programs like Medicaid.

Only 27% opposed it, and 13% weren't sure.

Medicaid's own improper payment rate jumped from 5.09% to 6.12% in a single year, worth $37.39 billion in fiscal 2025 alone.

Republicans backed the idea at 74%.

Independents came in at 57%.

Democrats landed at 48%, more supporters than opponents even in the party that spent the last year screaming about DOGE.

That's not a partisan split. That's a stampede.

Voters are tired of watching their money vanish into a system nobody polices.

The Fraud Numbers Behind This Poll Are Worse Than Anyone Admits

The numbers behind the poll explain exactly why voters feel this way.

Federal agencies reported $186 billion in improper payments for fiscal 2025, and 82% of that was money the government just handed out and never should have.

The Government Accountability Office puts the real annual fraud losses as high as $521 billion.

House Republicans on the DOGE subcommittee, led by Rep. Marjorie Taylor Greene, put a number on the damage going back further: $2.7 trillion in improper payments since 2003.

Fraud investigators told that same hearing the federal fraud rate runs near 20%, compared to roughly 3% in the private sector.

Meanwhile the government spends less than 1/20 of 1% of the entire SNAP budget on catching the people stealing from it.

Haywood Talcove, CEO of LexisNexis Risk Solutions, told the committee there's "no excuse for the government to lag" when private industry already verifies identities on the front end to stop fraud before a check ever goes out.

This isn't new. Pandemic unemployment programs alone lost between $100 billion and $135 billion to fraudsters, according to the GAO, because states handed out checks without basic identity checks.

Treasury's Do Not Pay system caught 363,386 bad payments worth $2.6 billion in a single fiscal year once agencies actually turned the tool on.

The Congressional Research Service found agencies lost $765.8 billion to overpayments tied to bad data access between fiscal 2021 and 2024.

Roughly 73% of that, $556.6 billion, came from agencies failing to check data they already had sitting in front of them.

Even Cato Institute budget director Romina Boccia, no cheerleader for federal spending programs, says AI "cannot compensate for weak eligibility verification, incomplete audits, poor data sharing, or a lack of accountability when agencies and contractors make repeated errors."

Translation: the tool was never the problem. The agencies that refused to use what they already had were.

That's the pattern. The technology works when someone bothers to use it. The fraud explodes when nobody does.

Democratic Voters Are Quietly Breaking With Their Party's Leadership

Democrats spent the entire year mocking DOGE as a cruel stunt aimed at hurting poor people.

Now their own voters are telling pollsters they want the AI turned loose on fraud anyway.

That's not a coincidence. That's a base in open revolt against its leadership's talking points.

Watch what happens next in Congress.

Republicans on the DOGE subcommittee already have the receipts, the hearing testimony, and now a bipartisan poll to back a permanent AI fraud-screening mandate across every federal agency.

Democrats who vote against it will have to explain to their voters why they blocked something 48% of their party wanted.

The bureaucracy has hidden behind "complexity" for decades to avoid accountability.

A machine that flags a stolen Social Security number in seconds doesn't care about complexity.

That's exactly why the people cashing in on the current mess are so quiet about this poll.

Hell, with all the hellacious slop AI – which is the name of a city in the Bible that translates to “heap of ruin” – is unleashing on society, the least it could do is stop fraud.

Sources:

• Casey Harper, "Bipartisan majority of Americans wants AI aimed at government fraud, poll finds," Just The News, September 2026.
• U.S. Government Accountability Office, "Fraud and Improper Payments: Data Quality and a Skilled Workforce Are Essential for Realizing Artificial Intelligence's Benefits," GAO-26-108850.
• U.S. Government Accountability Office, "Unemployment Insurance: Estimated Amount of Fraud During Pandemic Likely Between $100 Billion and $135 Billion," GAO-23-106696.
• Congressional Research Service, data-access-related improper payments analysis, fiscal years 2021-2024, cited via Just The News, September 2026.
• Romina Boccia, "Letter to the Editor: Fighting Fraud Won't Fix the Federal Budget," Cato Institute.
• Centers for Medicare & Medicaid Services, "Fiscal Year 2025 Improper Payments Fact Sheet," CMS.gov, January 2026.
• House Committee on Oversight and Government Reform, "Hearing Wrap Up: DOGE Subcommittee's First Hearing Uncovers Billions Lost to Fraud and Improper Payments, Launches War on Waste."
• U.S. Department of the Treasury, Do Not Pay Business Center Fiscal Year 2024 Report.