New York Gaming Commission Chair Brian O’Dwyer warned the state’s mobile sportsbooks that they should create a way to share information on at-risk customers or face a regulatory solution of the state’s design. He said an intervention regime would have limited value if a customer flagged by one operator could simply open an account elsewhere.
At the commission’s Oct. 9 meeting, O’Dwyer said the industry was “on notice” to solve the problem, adding that a patron could exhaust their welcome with one of New York’s nine mobile sports-wagering operators and move through the rest. He did not specify what form a shared data system should take.
The warning followed the commission’s unanimous 6-0 vote to advance draft rules governing artificial intelligence in wagering and safeguards for at-risk mobile sports-betting customers. As we reported on Oct. 9, those proposals would require operators to identify potentially vulnerable bettors and limit personalised betting offers.
The proposed customer-intervention rule, modelled on a New Jersey framework, would set escalating responses when bettors meet problem-gambling indicators. A March draft listed 13 potential triggers, including deposits exceeding $10,000 in 24 hours and account turnover of more than $1 million over 90 days.
The measures would initially direct affected customers to deposit, loss and time limits. They could then be required to watch an educational video before continuing to wager, followed by direct operator contact concerning their gambling. The draft also would prevent a customer from reversing a withdrawal request for three days.
The proposal includes additional financial-vulnerability checks in certain circumstances. One example is a weekly increase of at least 50% in time spent on a betting application compared with the preceding two-week period.
A separate proposal, Part 5404, would bar operators and a broader group of New York gambling licensees from using AI or automated systems that process individual customer data to personalise promotions, bonuses, odds, limits, messages or product recommendations. It would also prohibit such systems from suggesting bets or stake sizes, or from being used to increase wagering, engagement or revenue.
The AI restrictions would not apply to tools for fraud prevention, identity verification, anti-money-laundering checks or responsible-gambling functions designed to reduce betting. Uniform offers for all eligible customers would remain permitted. Licensees would have to maintain inventories of relevant automated systems, preserve logs for at least seven years, allow commission audits and give regulators 30 days’ notice before material system changes.
The draft rules were first circulated for industry feedback in March amid Governor Kathy Hochul’s effort to address gambling addiction. They followed ProPublica reporting that examined how DraftKings and other betting applications treated customers displaying signs of problematic play; DraftKings said it already followed New Jersey’s threshold framework.
The Responsible Online Gaming Association, formed by seven large operators in 2024, had separately pledged to share information about problem gamblers through a planned independent clearinghouse or database. Its members represented more than 85% of the legal U.S. online betting market, although the details of that proposed system were not clear.
Neither New York proposal is in effect. Both entered a 60-day public-comment period and are due to return to the commission, with no final-vote date set.
