Rogue brokers exploiting ACA enrollments
- 183,553 complaints of unauthorized enrollments were logged by CMS Jan–Aug 2024; CMS resolved 99.75% of them, with an average ~52-day cycle time.
- CMS also reported ~50,000 unauthorized enrollments and ~40,000 unauthorized plan switches in Q1 2024 alone.
- Enforcement accelerated: 850 brokers were suspended for suspected fraud/abuse tied to unauthorized enrollments or switches (Jun–Oct 2024).
- Real-world harm: Florida brokers targeted vulnerable people, including those experiencing homelessness, to generate commissions via zero-premium plans (multiple investigations and state reports).
- Criminal accountability: a Florida insurance executive pled guilty in April 2025 to an ACA enrollment scheme that triggered $133.9M in federal subsidies for ineligible enrollees.
Ineligible applicants gaming eligibility (identity/income misreporting)
- CMS’s broker actions sit alongside persistent eligibility fraud risks: KFF’s 2025 brief documents widespread unauthorized enrollments/switching and summarizes policy moves to tighten verification.
- Public debate over the scale of income misreporting exists (e.g., Paragon Institute estimates millions of questionable enrollments; other officials put the figure closer to 1–1.5M), but all sides agree verification needs strengthening.
- Program-wide leakage context: Medicare FFS improper payments were $31.7B (7.66%) in FY2024; Medicare Advantage $19.07B (5.61%); Part D $3.58B (3.70%). These are not all “fraud,” but they quantify the verification gap that criminals exploit.
- Medicaid improper payment rate was ~5.09% in 2024, underscoring the need for up-front eligibility controls.
Organized crime & money laundering targeting Medicare/Medicaid
- Largest takedown in DOJ history (June 30, 2025): 324 defendants, $14.6B in alleged fraud; $245M seized; CMS prevented >$4B from being paid; 205 providers had billing privileges suspended/revoked pre-takedown.
- Operation “Gold Rush”: a transnational organization acquired dozens of DME suppliers, used >1,000,000 stolen identities across all 50 states, and submitted $10.6B in false Medicare claims—explicit money-laundering via crypto/shells.
- Local lens: a CT provider acquired in 2023 submitted $7M in false claims within months; proceeds were laundered through shells and crypto—one of many U.S. entities used by the same network.
“Prevent-and-Prove” with CITIZ3N Verify — what we measure (deployment KPIs)
To end pay-and-chase, we make front-door integrity measurable at enrollment/renewal.
A. Identity assurance (stop impersonation; stop straw enrollments)
- Goal KPI: ≥98% automated identity proofing on first pass; <1s risk-score latency; <0.1% false-accept rate for synthetic identities.
- Evidence baseline: DOJ/OIG show stolen IDs fueling multi-billion schemes; CMS already blocked unauthorized broker changes as of July 19, 2024—CITIZ3N adds continuous identity binding at account, device, and broker levels.
B. Broker conduct analytics (find rogue patterns early)
- Goal KPI: ≥60% reduction in unauthorized enrollment/switch complaints in pilot states within 2 open-enrollment cycles; TTR (time-to-remedy) <14 days for flagged cases via three-way verification calls and digital attestations.
- Evidence baseline: 183,553 unauthorized-enrollment complaints in 8 months 2024; 850 broker suspensions in 2024 show where analytics should focus.
C. Employment & income verification (shut down ineligible subsidies)
- Goal KPI: ≥85% of income declarations cross-verified against wage/payroll/benefit streams in seconds; ≥25% reduction in post-enrollment income discrepancies year-over-year; <2% manual review rate.
- Evidence baseline: Disputed scale aside, credible reporting confirms widespread income misreporting exposure—CITIZ3N closes the loop with layered payroll/tax/wage checks before APTC flows.
D. Asset & entity risk (follow the money, not just the member)
- Goal KPI: >90% of high-risk entity relationships (shell owners, straw acquisitions) flagged pre-payment; ≥70% precision on first-pass beneficial-ownership matching.
- Evidence baseline: “Gold Rush” shows straw ownership and rapid DME acquisition as the attack vector; we fuse corporate registries, sanctions/PEP, bank-like AML heuristics to pre-empt payment.
E. Program outcomes (what leaders care about)
- Prevented payouts: track $ prevented (claims denied/held) per quarter—aligned to DOJ/CMS reporting; target >5:1 ROI in year one (prevention value / platform cost).
- Member protection: >90% restoration accuracy for victims of unauthorized switching/enrollment within 7 days (forms, tax corrections, benefit restoration).
- Auditability: evidence-grade case files (event timeline, data provenance, model versioning) to satisfy OIG/AG reviews with 100% replayability.
How it works in practice (fast start)
Broker risk controls (day 1)
- Enforce “known-agent only” change rules with device fingerprint + consumer OTP + voiceprint where allowed.
- Auto-flag outlier behaviors (e.g., 1 agent → many counties; midnight switches; seasonal bursts).
- KPI to watch: complaint rate per 10k enrollments. (CMS tracked 183,553 in 8 months—your before/after line.)
Identity + employment/income (first 30–60 days)
- Bind person→account→device; verify SSN/DoB + wage streams before APTC or Medicaid activation.
- KPI to watch: % of APTC determinations with verified wages; discrepancy rate at redetermination.
Asset/ownership graph (first 90 days)
- Beneficial-ownership graph + shell heuristics before payments to suppliers or MA plans; deny/suspend on adverse risk.
- KPI to watch: $ held/denied pre-payment (CMS/DOJ show the benchmark: >$4B prevented pre-takedown in 2025).
Transparency & learning (ongoing)
- Monthly model drift reviews; quarterly public integrity reports (complaints, prevented $).
- KPI to watch: false-positive rate; mean time to clear (consumer-friendly).
Why this is different
We’re aligning to how DOJ/OIG/CMS already measure success: defendants charged, intended loss, $ prevented, assets seized, provider actions. That turns “AI against FWA” from a buzzword into accountable, public-sector KPIs.
We start where FWA starts—with brokers at the point of sale, identity at the point of enrollment, and ownership at the point of payment. Not months later in recovery.
Artificial intelligence isn’t just a tool for spotting anomalies after the fact—it’s redefining how program integrity is measured and enforced. By embedding AI-driven identity proofing, broker conduct analytics, income verification, and ownership graphing directly into enrollment and payment workflows, agencies can move fraud detection upstream where the risks actually originate. This means preventing billions in improper payouts before they occur, protecting consumers from exploitation, and providing regulators with evidence-grade audit trails that stand up in court. With solutions like CITIZ3N Verify, AI transforms the fight against fraud, waste, and abuse from a reactive chase into a proactive safeguard—delivering measurable outcomes, faster cycle times, and a sustainable return on investment in the integrity of our health programs.
Sources & evidence
- CMS on unauthorized enrollments/plan switches, resolutions & cycle time; and new broker-change blocking (July 19, 2024).
- CMS broker suspensions (850 in 2024).
- DOJ/OIG 2025 national takedown ($14.6B, 324 defendants, $245M seized, > $4B prevented, 205 providers acted against).
- “Operation Gold Rush” (multi-billion, stolen identities of >1M, laundering via crypto/shells).
- Localized laundering case (CT provider bought, $7M false claims in months).
- ACA broker abuse targeting homeless (investigations and state reporting).
- KFF brief on fraud/eligibility risks and tightening verification.
- Medicare improper payments FY2024 (FFS $31.7B, MA $19.07B, Part D $3.58B).
- Medicaid improper payments (2024 ~5.09%).
