Home
/ Insights
/ Articles
/ New Fraud Economics: How AI Is Reshaping Risk in Sports Betting

New Fraud Economics: How AI Is Reshaping Risk in Sports Betting

The U.S. legal sports betting market has grown at extraordinary speed since the Supreme Court opened the door in 2018.¹ With that growth has come a parallel surge in fraud similar to what’s happening to financial institutions.² This surge isn’t driven by lone actors or simple schemes, but by artificial intelligence (“AI”) operating on an industrial scale.

Generative AI has fundamentally changed the economics of fraud.³ Attacks that once required specialized technical skills and significant time investment can now be automated, scaled and executed at near-zero marginal cost. For regulators and operators navigating a rapidly maturing industry, understanding this shift is no longer optional – it is the central challenge of the next regulatory cycle.

A Converging Triple Threat

The AI fraud threat in sports betting operates across three distinct but interconnected domains: identity fabrication, content generation, and autonomous execution.

Identity fabrication has moved well beyond stolen Social Security numbers. Generative adversarial networks and diffusion models now produce photorealistic face images, video selfies, and voice clones capable of defeating facial recognition and liveness checks. AI generates forged driver's licenses and utility bills that pass automated document verification. The most sophisticated operators build layered synthetic personas over weeks, complete with credit history, social media presence, and consistent device fingerprints, specifically engineered to withstand enhanced due diligence. Synthetic identity fraud grew eightfold in 2025 alone.⁴

Content generation enables fraud at scale through channels that were previously labor-intensive. Large language models produce personalized phishing emails and SMS messages that precisely mimic the tone and branding of legitimate sportsbook communications. AI replicates the full user interface of licensed betting platforms, creating convincing clones where users unknowingly surrender their credentials. During major sporting events, deepfake endorsements, synthesized celebrity likenesses promoting fraudulent platforms, proliferate across social media and digital advertising.

Autonomous execution is where the volume becomes truly alarming. AI agents now autonomously create accounts, claim sign-up bonuses, and place bets across multiple platforms without human involvement. Agentic bot traffic on gaming sites rose 450% in 2025.⁵ Advanced bots replicate human cursor movements, typing cadence, and session behavior to evade behavioral biometric detection. Coordinated rings use AI to orchestrate multi-account, multi-platform wagering, placing complementary bets across sportsbooks to manipulate lines or launder funds in patterns that are nearly invisible when each operator sees only its own data.

These three domains do not operate in isolation. A mature fraud operation uses identity fabrication to create the accounts, content generation to build credibility and acquire victims, and autonomous execution to scale the entire operation. The result is a threat that is cheaper to deploy, harder to detect, and more damaging than anything the industry has previously faced.

Embedded at Every Layer

What makes AI fraud uniquely difficult to address is that it mirrors the legitimate technology stack. AI is now embedded in every transaction layer of the betting ecosystem, from onboarding through payouts, and the same capabilities that enable legitimate operations create the attack surfaces that sophisticated actors can exploit.

During normal customer onboarding, operators deploy automated document operational character recognition (“OCR”), facial recognition, liveness detection and device fingerprinting. Each has a corresponding attack vector: deepfake selfies bypass liveness checks, AI-forged documents pass OCR validation, synthetic identities evade exclusion lists, and global positioning system (“GPS”) spoofing defeats geolocation controls.

In trading and odds, the adoption curve has been steep. On one major network, 48% of bets are now traded by AI, up from 4% in 2022.⁶ Real-time line adjustment across thousands of markets per game is standard. But coordinated AI-driven betting can manipulate those same lines, while latency arbitrage and insider-informed algorithmic wagering exploit the speed and complexity of modern odds systems.

At the monitoring stage, behavioral biometrics and anomaly detection tools represent meaningful defenses, yet bots now mimic the exact behavioral signatures these systems are trained to flag. Bad actors distribute low-value bets below monitoring thresholds, rotate through synthetic identities, and exploit the gaps between operators’ siloed systems.

At payout, transaction pattern analysis, velocity checks, and cross-account linkage are designed to catch structuring and laundering. But AI-enabled structuring keeps withdrawals below reporting thresholds, multi-account cash-outs cycle through synthetic identities, and first-party fraud through chargeback abuse could potentially cost U.S. sportsbooks $2.8 billion in annual losses.⁷

Identity Theft as the Connective Tissue

Identity theft is the thread that ties these threats together. The sports betting industry is a uniquely attractive target for identity thieves because the normal transaction patterns – marked by rapid deposits, frequent wagers and fast cash-outs – provide natural cover for fraudulent activity. Unlike retail banking, where sudden activity on a dormant account raises immediate flags, the velocity and variability of betting behavior make stolen-identity transactions difficult to distinguish from legitimate play.

Fraudsters use stolen identities to open accounts, fund them from stolen financial accounts, place bets to simulate normal activity, and withdraw the funds to accounts they control. The operator absorbs the chargeback. Where platforms extend credit, identity thieves draw down lines based on the real person's financial standing and disappear. And with account takeover via credential stuffing and SIM swapping on the rise, even properly verified accounts remain at risk months after onboarding.

A Regulatory Imperative

Addressing these challenges requires action on several fronts.

First, regulators should mandate upgraded identity verification standards, including multimodal biometric checks, active liveness detection, and document forensics capable of identifying AI-generated forgeries.
Second, operators should be required to deploy AI-powered monitoring systems, including behavioral biometrics for continuous session analysis and real-time anomaly detection across betting patterns.
Third, and perhaps most critically, cross-platform data sharing must become the norm rather than the exception. When each operator sees only its own data, coordinated fraud rings operate in the gaps between systems.

Looking Ahead

The speed at which AI capabilities are advancing means that today's defenses will be insufficient tomorrow. Regulators who invest in AI literacy now – not to become technologists, but to ask the right questions and set the right standards – will be positioned to protect both consumers and the integrity of an industry that is still defining its regulatory maturity. The economics of fraud have changed. The economics of prevention must change with them.

Footnotes:

1: Mark Sherman, “How a 2018 Supreme Court decision paved the way for meteoric growth in legal sports betting,” AP News (October 24, 2025).

2: “FTC Testifies before the Joint Economic Committee on Agency’s Efforts to Combat Fraud,” Federal Trade Commission (March 25, 2026).

3: Antonio Gesteira, Martin Tupila, Veeral Gosalia, Richard Palmer, and Nicholas Newbury, “Understanding and Combating AI-Powered Fraud in Financial Ecosystems,” FTI Consulting (July 17, 2026), https://www.fticonsulting.com/insights/white-papers/understanding-combating-ai-powered-fraud-financial-ecosystems.

4: “Synthetic Identities and Agentic Bots Posing as Human Contribute to 8% Global Rise in Fraud Attacks,” LexisNexis Risk Solutions (March 26, 2026).

5: Ibid.

6: Charlotte Capewell, “AI Accounts for Nearly Half of Sports Bets on Kambi Network,” Yahoo! Sports (January 19, 2026).

7: Ori Snir, “First-Party Fraud Cost Online Gaming Operators $2.8 Billion in 2024, New Report Finds,” Socure (January 30, 2025).