.svg)
AI chatbots are already driving down problem gambling risks and freeing up compliance staff, according to an advocate from one of the world’s tech giants, but other experts warn of risks when regulators come calling.
The use of AI is now a daily reality in gambling and the industry is grappling with what its early deployment means for businesses.
One potential use is to power chatbots that act as a virtual buddy for customers, including encouraging them to take a break when at risk of harm.
Jason Clairmont, the principal evangelist for betting and gaming at Amazon Web Services, claims to have seen data from an operator in the US that showed a 40 percent increase in players opting to take gambling breaks when the nudge comes from an AI, rather than a real human being.
He described it as a “softer experience”, adding, “you’re already interacting with a bot that you think of as a friend.”
An AI can use a player’s age and other features, combined with what it learns from interactions, to communicate in a way tailored to be more impactful, Clairmont said.
Similar tools are also being rolled out within compliance departments at various gambling companies, Clairmont said.
The tech helps shortcut work that would otherwise be done manually, but its outputs still need to be verified, he said.
“It’s up to the human in the loop to come up with the right answer at the end of the day,” he cautioned.
Clairmont spoke alongside other industry experts last week at a London conference organised by law firm CMS.
Adam Rivers, who oversees betting and gaming at consulting firm Alvarez & Marsal, noted that using AI in any stage of compliance creates risks when regulators knock on your door.
“The big thing is explainability. If you’re going to make a decision that’s going to impact a customer, you have to be able to fully justify it,” he said.
“From a regulator’s perspective, if you’re hauled in front of the Commission and you can’t show in your audit trail why you took a certain decision, it’s not excusable.”
“In this industry, you can’t ask for forgiveness,” added Clairmont, warning against taking the approach like Uber or Amazon by innovating first and then worrying about compliance later.
Workers At Risk?
Content supplier G Games has already begun to test AI in a creative capacity, said the company’s director of operators, Jakub Szerszen.
In user testing, a fictional game character created by an AI tool outperformed those thought up by human designers.
He said this discovery “wasn’t taken well by our team”, but that the business was adapting to deploy staff where human input was most needed, including to realise ideas dreamt up by the computer fully.
Still, he said, AI has some way to travel before it can fully replace human workers in many roles.
For example, “you need to have someone behind [the AI] who understands how to write code. We are not at the state where you can let it write the full code,” he said.
The lure of countless man-hours saved by integrating AI into your business also presents traps of its own, the panel warned.
“Millions of dollars have been burned building things” that don’t work, said Clairmont, but that industry was largely “getting back to first principles” after an initial wave of excitement.
Looking To The Future
Looking into the future, the panel warned of a surge of new challenger firms, particularly in the supplier space, that will lean on AI to kickstart their business.
“It’s a 90 percent advantage to be a startup right now,” versus an established company, said Clairmont.
“There are companies coming on board that are able to spin up operations in weeks,” added Rivers.
For established firms dealing with unwieldy internal systems and disconnected business units, the job is much harder.
“Because of how complicated it feels, there is going to be a bit of inertia in decision making,” said Rivers.
Businesses will be afraid of moving too fast or too slow, or committing to the wrong provider. “It’s a scary crossroads to go through,” he said.
For their part, regulators are also gaining confidence in using and understanding AI, said Clairmont.
“In conversations, especially in the past 12 months, it’s pivoted pretty aggressively,” he said.
“A lot of them are using ChatGPT to make their lives easier, so they know how to use it, but there are some conflicts where the regulators are handcuffed by the regulations in their state.”
