UNLV report highlights gaps in igaming AI governance

Dig Deeper

Related stories

2025/06/02: Report highlights risk from illegal gambling platforms on Indian youth 2025/05/16: Study highlights risks of igaming in Laramie County, Wyoming 2025/11/03: Coinbase CEO’s on-air joke highlights flaw in prediction markets 2025/12/03: The United States Igaming Revenue Report – October 2025

The Backstory

Why the governance gap matters now

The University of Nevada, Las Vegas study lands at a moment when online gambling is scaling fast and regulatory expectations are tightening. The report’s finding that governance ranks lowest on an AI maturity index underscores a basic contradiction: operators are deploying machine learning across marketing, game operations and risk while the internal controls, defined roles and responsible-AI policies needed to manage those tools lag behind. That disconnect is not just a compliance issue. It goes to the heart of consumer protection, market integrity and trust — the foundations that will determine how far and how fast digital wagering can grow.

Recent industry trends magnify the stakes. U.S. online casinos and poker rooms posted record revenue across all seven legal states in October 2025, a clean sweep that highlights how demand is outrunning the control frameworks many companies still lack. As detailed in The United States Igaming Revenue Report for October 2025, Michigan, New Jersey and Pennsylvania each set new highs, contributing to a combined $907.4 million month. That growth strengthens the case for robust AI oversight — because the bigger and more automated the systems, the greater the downside when something goes wrong.

Scale without discipline raises operational risks

Operators are leaning on AI to personalize offers, detect fraud and streamline customer support. But without clear governance — accountable leaders, model inventories, testing protocols, bias checks and incident playbooks — those systems can drift, fail silently or create perverse incentives. The warning is not theoretical. Data on player behavior show American customers are depositing and wagering at higher levels than global peers, which amplifies exposure to missteps in targeting, affordability checks or self-exclusion controls.

In May 2025, the U.S. gambling customer base both grew and spent more than the global average, according to Optimove’s Gaming Pulse report. Average deposit amounts for U.S. players climbed 10% year over year to $604, while average monthly casino spend reached $8,259, over six times the global benchmark. Yet engagement and retention trailed other markets. That mix — higher spend with weaker stickiness — can push operators to lean harder on algorithmic segmentation and bonusing to maintain revenue, exactly where transparent, well-governed AI is most critical to avoid targeting vulnerable users or oversaturating big spenders.

Illegal markets exploit weak controls

Governance gaps do not exist in a vacuum. Bad actors actively probe them. In India, where fragmented oversight leaves gray zones, illegal platforms have siphoned an estimated near-$100 billion in annual deposits and attracted billions of visits, often by leveraging private channels and lax identity checks. The recent analysis by Consumer Unity & Trust Society, covered in a report on the risks to Indian youth, details how these sites bypass basic safeguards and target younger users. While the U.S. legal market is more structured, the tactics overlap: frictionless onboarding, aggressive promotions and the ability to relaunch under new domains when blocked. If regulated operators lag on governance — including verification, affordability modeling and real-time anomaly detection — they risk narrowing the practical gap with illicit competitors and weakening the argument for channeling players into licensed environments.

AI is central to that contest. Well-governed systems can strengthen age checks, flag proxy use, detect synthetic identities and surface suspicious transaction patterns. Poorly governed systems can do the opposite — scaling aggressive acquisition without adequate filters, or creating opaque decisions regulators cannot easily audit. The UNLV findings that few companies have embedded responsible-AI policies and that hiring for governance roles remains sparse should worry both operators and policymakers who want licensed offerings to be clearly safer and more trustworthy than their illegal counterparts.

Local data puts human costs in sharper focus

Community-level research illustrates how digital convenience changes risk. A study shared by the Cheyenne Police Department found that online gamblers in Laramie County, Wyoming, play longer, exceed budgets more often and report higher rates of anxiety and strained relationships than other gamblers. As detailed in the Laramie County analysis, young adults are especially vulnerable, and many at-risk users are reluctant to seek help or unaware of local resources. The report points to basic “bumpers” — spending limits, time reminders, self-exclusion — that online platforms can implement, but the effectiveness of those tools depends on how they are designed, surfaced and enforced.

This is where governance meets product. AI can preempt harm by predicting risky play patterns and escalating interventions, but only if models are validated, thresholds are conservative, overrides are restricted and outcomes are audited. Without those controls, harm-minimization can become a compliance checkbox rather than a functioning system. The Wyoming findings echo a broader policy push: strengthen consumer protection before expansion accelerates further and before illicit operators widen their reach.

Market integrity and the lessons of manipulability

Robust governance also protects market integrity when real-time data, incentives and public narratives intersect. The ease with which Coinbase CEO Brian Armstrong triggered payouts on “mention markets” during an earnings call illustrates how quickly self-referential systems can be gamed. The episode, detailed in coverage of prediction market vulnerabilities, shows that when a market’s outcome depends on the actions or words of a small set of actors, manipulation is not just possible, it is almost trivial. While these were not sportsbook markets, the principle travels: products that blend public signals, influencer dynamics and rapid settlement are fragile without strong controls on conflicts, surveillance for manipulation and clear rules on employee conduct.

For betting operators experimenting with new formats — micro-events, peer-to-peer markets, creator-driven props — governance needs to front-run innovation. That includes defining red lines on markets “readily susceptible to manipulation,” enhancing surveillance models and logging interventions. Regulators will ask for that evidence when novel products emerge. Firms that invest early in responsible-AI pipelines and controls will be better positioned to respond.

The road ahead: from pilots to policy

The industry is not starting from zero. Most operators now use AI in some capacity, and academic interest is rising in applications such as detecting at-risk play. The gap is in codifying those capabilities into accountable structures: appointing leaders for AI risk, inventorying models, instituting bias and performance testing, embedding responsible-use policies, and ensuring that harm-minimization systems are measurable and independently auditable. As the revenue base expands and customer spend climbs, the cost of delay increases.

Expect regulators to move from principles to proofs. Where companies can demonstrate that AI systems are documented, monitored and governed — and that consumer outcomes improve as a result — they will gain credibility to innovate. Where they cannot, enforcement, product restrictions or reputational hits are likely. In competitive terms, governance is shifting from a back-office chore to a strategic differentiator: it will decide who can safely scale and who will be forced to pause.