-Artificial intelligence, or AI, relies on data centers, and its use is growing quickly.

According to a recent report from consulting firm McKinsey & Company, 88% of organizations they surveyed reported using AI in at least one business function.

But what about AI's use in gaming?

A new report from UNLV's International Gaming Institute looks to track AI's progress in the global gambling industry on an annual basis.

We recently spoke with Kasra Ghaharian, editor-in-chief of the inaugural report which found a disconnect between gaming companies and the entities that regulate them.

(Kasra Ghaharian) We did a survey of over 80 gambling companies across the world.

We also surveyed over 110 regulators around the world as well, so giving us this kind of dual insights into what's going on with AI in the industry.

What we found was that regulators expressed they had very little visibility into how AI is being used amongst their licensees, which actually tracked in the data.

So we asked the industry, you know, where is the most significant activity in terms of your AI use?

They said things like back of house, like product innovation, technology, and security.

Customer-facing functions were actually, you know, they were further down the line.

Yet when we asked regulators the same question, they thought customer-facing functions was where all the activity was.

So a definite disconnect there in terms of where regulators think it's being used and where it's actually being used.

-How would you describe that?

Is that alarming or maybe par for the course, given the-- given how new AI is?

-I think the main issue with this is that while regulators express they have little visibility into how it's being used, they have appetite for regulating it.

They think that there needs to be gambling-specific AI regulations.

A lot of them think that these regulations need to be stricter than or on par with broader AI regulations.

So the question is, you know, while there's low awareness of how AI is being used on the ground, there's high conviction in terms of the need to govern it, which I think is a dangerous mix, because if you don't understand why it's being used but you want to regulate it, you might end up with regulations that don't really match what we need to regulate, if that makes sense.

So I mean we've seen some kind of activity around regulatory action around AI.

We're starting to see signals that that might increase in the future.

We did a regulatory scan with this company called Vixio.

So we looked at all of these kind of regulatory action that was going on.

What we found was that activity is kind of concentrated in terms of encouraging or in places mandating the use of AI, automated methods, data for monitoring and detecting at-risk players.

So that was the main area of concentration.

We saw less so in terms of specifically regulating AI.

I think there was only a couple of-- There was only a couple of places we saw specific language around AI, specifically.

Massachusetts has something like prohibiting the use of AI to make platforms more addictive.

New York, just last month, I think, they have proposed language around banning the use of AI for marketing and targeting players.

So we're starting to see signals that there might be more AI-specific regulation in gambling.

-You talked about the regulation currently happening, and I think there are a couple states that were mentioned--Colorado and New Jersey, North Carolina also--that are using it to seek out problem gambling.

When someone is maybe gambling too much, they can identify that and then perhaps step in.

I was surprised not to see Nevada on that list, considering Nevada is looked at as the premier regulator of gaming.

Any insight into where Nevada stands in all this?

-Yeah.

I think you're correct.

I think, you know, Nevada, you know, the Gaming Control Board, the regulator, they're called upon from jurisdictions around the world because they're seen as a leader in this area.

I wouldn't say the lack of any kind of language or action around AI specifically for responsible gambling purposes is something bad.

What I see it as is maybe a philosophy.

So I think maybe, you know, there's two schools of thought, right?

There's on the one hand you could have regulatory action and language specifically calling out the use of AI for responsible gambling.

On the other hand, you could take an outcomes-based approach.

So with that I mean, okay, you're concerned about the outcome in terms of player harms.

So that's what you regulate.

You don't care how your licensees kind of get there to protecting the players, whether it's AI or whether it's some other method, you're concerned with that outcome being met.

And I think that's quite a good approach, because it's maybe a bit more-- And maybe age is better, because technology changes a lot.

So if you're regulating a specific technology, that technology could change or there could be a new technology that comes around.

If you take an outcome-based approach, whatever technology comes around, your outcome-based approach is maybe, maybe a bit more adaptable to those changes.

-What are some of the, you call them "AI incidents" in the report, that stand out to you when there is harm that comes from the use of AI in gaming?

What did you find?

-Yeah.

So we, we thought it would be nice to track any AI-related incidents.

We're doing the report on an annual basis, so we thought year over year we might be able to track these incidents, see if there's an increase in incidents, see if there's an increase in any specific types of incidents.

So we looked at these publicly-available databases.

For example, there's one called the AI Incident Database.

I think it's run by a nonprofit.

But you can look in these databases, and you can filter for kind of like gambling specific ones.

So that's what we did.

We found like most of the incidents that had been logged on these databases were around deep fakes, so using deep fakes to promote illegal gambling websites in various jurisdictions around the world.

So you know, people creating, you know, your favorite celebrity promoting some gambling website to try and get people to, you know, gamble on them.

So that was one of the emerging uses of AI.

-There was another one out of Reno, though, and that was with facial recognition.

And I think the-- it misidentified someone and led to an arrest?

-Yeah.

You know, and this, this-- I think that incident-- Computer vision is only just starting to be used, I think.

I don't think we're seeing kind of prevalent use of the technology as much as it could be.

There's various reasons for that, but I think that does highlight, you know, the need for governance.

We're seeing that the industry is maybe training a bit in terms of their governance, practices, and guardrails.

So this could be why we saw an incident like this happen.

But you know, as I said, these are public repositories.

They rely on, you know, stakeholders, operators, journalists, the public logging these incidents, so that could be one of many.

-Many, right.

Because the casinos themselves are probably not going to report themselves.

-Yeah.

-What about the gaming companies' ability to regulate themselves, their governance?

How prevalent is that of AI in their companies?

-Yeah.

This was probably like the key finding of the report, I think.

So, as I said, we surveyed over 80 gambling companies.

These spanned land-based operators, online operators, suppliers as well, you know, online and land-based suppliers.

And we asked them various different questions, but we created this AI maturity index across four dimensions.

So we asked them about their AI strategy, you know, do they have one?

How developed is it?

Is it actively guiding their initiatives?

We also asked questions about their infrastructure, their AI expertise, and we also asked them about their governance.

Do they have responsible AI or AI governance procedures in place?

So those were the four dimensions: strategy, infrastructure, expertise, and governance.

And we found a big gap in terms of how they scored on AI strategy and how they scored on governance.

I think it was about a 30-point gap, actually.

So what that tells us is that these industry players have high ambitions.

They want to adopt the technology to help them with cost efficiency, revenue generation, things like this, yet they don't have the same appetite to put the guardrails in place so that the technology is governed in a responsible manner.

-How do you even get there then?

And is this not hiring the right people within the AI sphere?

-Yeah, it could be.

We asked that question, actually.

So, you know, there's very few companies that have dedicated AI governance roles, which is dissimilar to what we see in other sectors.

For example, financial services, you see some of these companies--JP Morgan, Capital One--they have individuals dedicated to AI governance.

They even have AI governance teams.

We're not seeing that in the gaming industry yet.

I think that might change.

You know, the gaming industry has spent decades kind of establishing frameworks for things like anti-money laundering, responsible gambling, and they've done an excellent job and they've gained trust from regulators and the public from that activity.

So I think the same thing will probably happen with AI.

I think we'll start to see the responsible AI VPs or departments emerge in the industry.

-All right.

Kasra Ghaharian with UNLV, thank you so much for joining Nevada Week.

-Thank you.

My pleasure.