Massachusetts Gaming Commission to examine gambling operators’ AI use

Dig Deeper

The Backstory

AI scrutiny lands in an already assertive market

The Massachusetts Gaming Commission’s review of how gambling operators use artificial intelligence is the latest test of a regulatory model that has taken a comparatively aggressive posture since online sports betting launched in the state. The inquiry follows a New York Times report alleging that DraftKings used machine learning to identify customers more likely to respond to promotions by continuing to gamble and lose money. DraftKings has disputed that account, saying it complies with Massachusetts rules and does not use AI to target customers based on losses or signs of problem gambling.

The stakes are heightened because Massachusetts rules already contemplate the risk that automated systems can turn customer data into a tool for harmful gambling behavior. Operators are barred from using customer information for promotions generated by automated systems, including AI, when those systems are known or reasonably expected to increase addiction risks. They also must report to the commission on how customer behavior is analyzed for promotional and responsible gambling purposes.

That framework gives the commission a clearer basis for asking questions than regulators in many other states. But it also creates a burden: If AI is embedded across promotional spending, customer segmentation, risk modeling and responsible gambling systems, the agency must determine whether existing reporting is enough to distinguish ordinary personalization from conduct that exploits vulnerability.

DraftKings allegations sharpen the policy question

The immediate trigger was the New York Times investigation alleging DraftKings used AI to target high-loss gamblers. According to that reporting, DraftKings began developing a machine learning model in 2023 that used betting frequency, account balances, historical losses and other customer data to estimate how much additional gambling losses a user might generate after receiving promotional offers.

The report said former employees raised concerns that the system could direct free bets and bonuses toward people more likely to keep gambling and lose. It also said the company considered, then canceled, projects that would have used similar technology to identify customers at risk of gambling problems. DraftKings said its promotions are directed toward customers with sustained engagement, not customers selected because of losses, and said predictive tools for identifying problem gambling lacked sufficient evidence of effectiveness.

The distinction is central to the Massachusetts review. Sports betting operators routinely use analytics to allocate promotions, manage customer acquisition costs and measure lifetime value. Regulators are now being asked to decide when those commercial practices cross into conduct that worsens addiction risk. If an algorithm predicts promotional profitability because a customer is likely to wager more, lose more or show signs of compulsion, the same data that improves margins may become evidence of regulatory risk.

That tension is especially acute for publicly traded operators under pressure to improve profitability. The earlier report noted that DraftKings has said AI-driven analytics helped improve margins on promotion-driven sports wagers. For regulators, that raises a practical enforcement challenge: models may not need to be explicitly designed around problem gambling indicators to produce outcomes that disproportionately affect vulnerable customers.

Credit card case showed enforcement appetite

The commission has already demonstrated a willingness to impose large penalties when it finds sports betting operators failed to meet Massachusetts requirements. In July, the agency imposed a $450,000 fine on DraftKings for accepting credit card wagers, the largest sports betting fine issued by the regulator against an operator, according to NBC Boston.

Massachusetts banned credit cards as a form of sports wagering payment when legal betting began in 2023. The commission found DraftKings allowed 1,160 impermissible wagers funded by 242 credit card deposits from 218 customers, with total handle of $83,667.92. DraftKings was ordered to repay affected customers and a third-party auditor was directed to determine whether additional credit card funds were used during the early launch period.

That case matters for the AI review because it shows the commission is prepared to treat compliance breakdowns as material even when the dollar amount involved is modest relative to an operator’s business. It also reflects the state’s view that payment methods, marketing and responsible gambling rules are not technicalities. They are part of the consumer protection structure lawmakers and regulators put around online betting when they authorized the market.

The AI issue could be more difficult to prove than the credit card matter. Payment violations leave transaction records. Algorithmic targeting requires access to model design, training data, business objectives, testing results and internal controls. The commission’s review will likely turn on whether operators can explain how automated tools are built, what variables they use, how promotions are approved and whether responsible gambling safeguards are integrated into commercial systems.

Advertising gaps raised earlier warning signs

Massachusetts’ concern with gambling marketing predates the AI allegations. A state audit released in August criticized the commission for failing to review online sports betting advertisements before they reached the public. The audit found shortcomings in betting ad oversight, including 17 advertisements that lacked the state’s problem-gambling helpline and 51 instances in which marketing materials were sent to people younger than 21 or people with gambling addictions.

The findings were damaging because they covered the early period after legal sports betting began, when operators were spending heavily to acquire customers and shape consumer habits. Massachusetts regulations prohibit marketing to minors and bar content designed to appeal to them. Auditors also found gaps in training documentation for casino employees assigned to assist at-risk gamblers.

The commission said it was implementing recommendations and had hired an independent auditor to assess operator compliance. But the audit framed a broader problem that now applies to AI: regulators may not be able to rely on after-the-fact review if marketing technology moves faster than compliance systems. Promotional campaigns can be generated, segmented and adjusted rapidly. AI tools may accelerate that process further, making it harder to identify improper targeting before consumers are exposed.

Public health advocates have argued that gambling products require proactive oversight because harms may compound quickly for at-risk customers. That argument is likely to influence how the commission evaluates automated decision-making. If AI determines who receives offers, when they are sent and how incentives are calibrated, the commission may ask whether operators need prior approval, independent testing or more detailed disclosures.

Unlicensed betting fight broadens the consumer-protection frame

The AI review also comes as Massachusetts officials challenge new products that blur the boundary between financial markets, gaming and sports wagering. Attorney General Andrea Joy Campbell filed a lawsuit against Kalshi over alleged illegal sports wagering operations, arguing that the prediction market was offering sports bets through event contracts without a state sports wagering license.

The complaint said Kalshi’s sports contracts, including moneyline, point spread and over-under products, closely resembled wagers offered by licensed sportsbooks. The attorney general alleged the company bypassed consumer protections required of licensed operators, including age restrictions, responsible gambling safeguards, compliant deposit limits and state oversight. Kalshi is regulated by the Commodity Futures Trading Commission, creating a jurisdictional dispute over whether sports event contracts are financial products or gambling products.

That case is relevant because it shows Massachusetts officials are focused not only on licensed operators but also on emerging business models that may avoid the state’s gambling rules. The commission and attorney general have emphasized age limits, addiction risk, advertising standards, taxes and player protections as nonnegotiable features of the legal market.

AI fits into the same policy arc. Whether the product is a sportsbook, lottery platform or prediction market, state officials are increasingly concerned that digital gambling can scale faster than oversight. The core question is whether consumer protections designed for conventional wagering remain effective when platforms use real-time data, automated predictions and personalized incentives.

Expansion increases the need for guardrails

Massachusetts is not retreating from digital gambling. The state lottery recently launched its first online lottery program through Aristocrat Interactive’s Massachusetts ilottery platform, a turnkey system that includes player account management, draw games, business intelligence, responsible gameplay tools, marketing services, data analytics, compliance and customer support.

The lottery rollout underscores why the AI debate matters beyond DraftKings. Digital gambling businesses depend on data-rich systems that personalize customer experiences and manage engagement. Those tools can support compliance and responsible gambling if used to identify risk, enforce limits and intervene early. They can also increase the intensity and precision of marketing if incentives are optimized mainly for revenue.

Massachusetts now faces the task of separating acceptable innovation from harmful targeting. The commission has not previously taken enforcement action specifically tied to AI, but its review could establish expectations for model governance, audit trails and how operators document the relationship between promotional analytics and responsible gambling obligations.

The outcome will be watched beyond the state. Maryland and federal policymakers are also considering AI-related gambling rules, while operators continue to expand automated marketing and risk systems. If Massachusetts concludes that current disclosures are inadequate, it could push the industry toward more explicit AI controls. If it finds no violation, the review may still clarify what operators must be prepared to show when algorithms shape who gets an offer and why.