.svg)
Regulators need to take a risk-based approach to navigate the dangers and opportunities presented by artificial intelligence (AI) in the gambling industry and avoid banning its use, according to a panel of AI specialists.
Speaking at the International Association of Gaming Regulators (IAGR) joint conference with the International Masters of Gaming Law (IMGL) in Rome on Tuesday (October 22), Mark Pace, vice president of the International Gaming Standards Association (IGSA), warned against the US legislature's attempts to ban the use of AI in sports betting through the SAFE Bet Act.
The SAFE Bet Act proposes prohibiting AI from tracking individual players’ gambling habits, creating individualised offers and promotions to customers, or using AI to create gambling products, such as microbets, something Pace believes will not work.
“Don’t ban it, understand what it is and what it consumes. AI is a hungry beast and you must understand the data that an AI application will consume.
“As it relates to using data to understand what people are doing and tailor offers to them, that is exactly what marketing is. Trying to say using technology [for marketing] is banned or prohibited makes zero sense,” Pace said.
Sukriti Verma, the head of analysis at the UK Department for Digital, Media, Culture & Sport, which oversees gambling, said the UK government’s perspective is “pro-innovation and growth”.
“Whenever there is a market failure, only then we are positioned to intervene. There are lots of risks with AI, but there are lots of opportunities too. Our job is to manage the risks, strike a good balance, and make sure it is all citizen-centric, proportional and pro-innovation. We shouldn't be banning anything,” Verma said.
When questioned about the incoming EU AI Act, Ian Gauci, managing director of law firm GTG, explained it is “well-studied” and a “different beast altogether” from the SAFE Bet Act.
“The EU AI Act is a horizontal piece of legislation. It is based on a risk approach. It mandates a risk approach similar to AML or GDPR. In a nutshell, there are banned uses of AI in it, but it is based on risk,” said Gauci.
Regulators should also be wary of the speed at which things are evolving in the AI space and ensure they are ahead of the curve, according to Verma, who warned that the potential bias in data being fed into AI might percolate down to the user, resulting in the unfair treatment of consumers.
“There is a complexity around the overlapping legislations. We have the EU AI Act, the UK government is also preparing itself for another AI Act. We need to make sure it's not too complex, it should be easy to understand, explainable, and easy to follow for the operators. All of those things are big challenges to achieve,” Verma said.
To overcome the complexity of AI regulation, Pace suggests regulators should take a step back and make sure gambling is fair, has integrity, and is done from a secure and safe perspective.
“Let's regulate it, make sure it is used in the right way. Keep it simple. Focus on the data. Understand what data you want the algorithms to use,” Pace said.
“Don’t worry about the AI engine, whether it's generative, or learning. Put together transparency requirements and responsibility requirements, and make sure you hold people accountable. Use a risk-based approach. Once you understand the data, you understand the risk of allowing it to be consumed by these engines.”
Separately, when asked by the audience, all the panellists agreed that the gambling industry should be more transparent about its use of AI.
