79% of iGaming Companies Use AI. But They Still Aren’t Using It Where It Matters Most

During the AffPapa Conference Madrid, our host Surya Palli spoke with Viktoriia Grygorenko, CEO of The Playa, for Episode 54 of iGaming Real Talk. And also the first episode where our host was asked questions as well. 

The conversation was based on The Playa's research into AI adoption across iGaming. More than 150 people started the survey and over 80 completed the full questionnaire.

The goal was simple.

Move away from the hype and understand what operators, suppliers and platforms are really doing with AI today.

Here’s what stood out.

There is a lot of AI talk, but much less clarity

This was the main reason The Playa started the research.

Everyone is talking about AI.

Companies say they are adopting it. Teams want to be ahead of the curve. New AI tools are appearing all the time.

But Viktoriia felt one thing was missing.

What is actually happening inside these companies?

As she put it:

“AI has a lot of noise. Everybody is talking about it, but nothing practical.”

The research looked at the practical side.

Who is already using AI?

Where are they using it?

Are they happy with the results?

And for companies that have not started yet, what is stopping them?

That made the conversation much more useful than another discussion about what AI might do one day.

The better question is what it is already doing now.

Big iGaming companies are moving faster than smaller ones

This was one of the biggest surprises in the research.

Normally, smaller companies are expected to move faster.

They have smaller teams. There are fewer approval layers. They can test things quickly.

Large companies usually take longer.

But according to Viktoriia, iGaming is showing a different picture.

“In iGaming actually the big companies are moving faster than the small and medium size.”

Larger operators often have bigger budgets and more resources.

They may also have started testing AI earlier.

That means some of the biggest businesses are already more mature across different AI use cases than smaller companies.

For smaller operators, this creates an interesting challenge.

Being smaller may make you flexible.

But flexibility only helps if you actually start experimenting.

The most common AI use case is not the most exciting one

During the conversation, Viktoriia turned the tables and started asking Surya questions from the research.

The first was simple.

Which business area has the highest AI adoption?

Surya guessed fraud detection.

The actual answer was much less dramatic.

Process automation.

And when you think about it, it makes sense.

Companies already have many repeated internal processes.

If the steps are clear, AI can help automate parts of them.

The result is also easier to measure.

Did it save time?

Did the team complete the work faster?

Did it reduce manual tasks?

There is less risk than giving AI control over something that directly affects player value or revenue.

Sometimes the best place to start with AI is not the most impressive use case.

It is the easiest problem to clearly define.

You need a good process before you can automate it

The discussion around automation also brought up another important point.

AI cannot fix a process that nobody understands.

As Surya pointed out during the conversation, if the process is not already defined, what exactly are you automating?

Teams still need to understand the job first.

They need to test it.

They need to know what a good result looks like.

Then automation becomes much easier.

This sounds simple, but it may explain why some companies struggle with AI projects.

Sometimes the AI is not the problem.

The process behind it is.

Player value prediction is much harder

At the other end of the research were areas such as bonus recommendations, player acquisition and LTV prediction.

These are much harder problems.

Predicting a player's lifetime value sounds like a very useful AI application.

But there is a reason adoption is still lower.

It needs good data.

Often, it needs many different data points working together.

And then there is another problem.

The operator needs to trust the prediction.

As Viktoriia explained:

“LTV prediction is tricky.”

An internal automation giving the wrong answer may create some extra work.

An LTV model giving the wrong answer could affect how much an operator spends to acquire or retain a player.

The bigger the business decision, the more important the quality of the data becomes.

The biggest AI barrier is not a lack of belief

Surya expected lack of knowledge or expertise to be the biggest thing stopping companies from adopting AI.

That was an important barrier.

But it was not number one.

The biggest barrier was competing priorities.

That tells us a lot about where AI currently sits inside many iGaming companies.

Teams already have roadmaps.

They have product updates.

Customer problems need to be solved.

Revenue targets need to be hit.

Then someone suggests an AI experiment where the result is still uncertain.

When something urgent happens, the experiment can easily move down the list.

It does not mean companies think AI is useless.

They simply have other problems that need attention today.

More than half say it is still too early to judge

Another interesting question was whether operators are happy with the AI they have already implemented.

Surya jokingly guessed they might be completely dissatisfied.

The answer was different.

More than half of the respondents said it was still too early to judge.

That may be one of the best descriptions of AI adoption in iGaming today.

Companies are testing.

They are learning.

Some use cases are working.

But many businesses still do not have enough data to say exactly what the long-term result will be.

There was one clear exception.

For simpler use cases such as process automation, satisfaction was much stronger.

Again, the pattern is quite clear.

The easier the problem is to define and measure, the easier it is to see whether AI is helping.

Nobody in the research said they don't believe in AI

Perhaps the most important finding came at the end of the conversation.

Viktoriia said 0% of respondents said they did not believe in the technology.

That does not mean everyone knows what to do with AI.

They clearly don't.

But the debate appears to have changed.

The question is becoming less about:

Should we use AI?

And more about:

Where should we use it first?

Companies are either already experimenting or planning to.

As Viktoriia put it:

“This is just a new norm.”

The results may still be early.

But companies are moving.

Starting small may be better than waiting for the perfect AI strategy

One thing became clear throughout this conversation.

You do not need to transform the entire business tomorrow.

In fact, that may be the wrong way to approach it.

Start with a clear process.

Find something measurable.

Test whether AI makes it better.

Then learn from it.

Because while many companies are still figuring AI out, waiting too long creates another problem.

Everyone else keeps learning while you stand still.

Viktoriia's message to operators that have not started yet was simple.

Start now.

Catching up becomes much harder when competitors already have months or years of learning behind them.

What stood out to us

1. There is still a big gap between talking about AI and actually implementing it.
2. Large iGaming companies are currently moving faster than many small and mid-sized businesses.
3. Process automation is one of the clearest places where AI is already being used.
4. Clear processes and good data need to come before useful automation.
5. Areas such as LTV prediction are harder because they depend on many data points and need a high level of trust.
6. Competing business priorities are a bigger barrier than disbelief in AI.
7. More than half of respondents believe it is still too early to judge their AI results.
8. Not one respondent said they did not believe in AI technology.

What stayed with us after this conversation was how normal AI is already becoming.

Not because every company has figured it out.

They haven't.

But almost everyone seems to agree that it will be part of the business.

The real difference may come from who learns how to use it well first.

You can also read The Playa's full “AI in iGaming: What the Industry Is Actually Doing” report here.

If you looked inside your company today, where is AI actually creating value and where is it still mostly talk?

As part of the 4th edition of The Real Roadshow 2026, Episode 54 was recorded at AffPapa Conference Madrid and is brought to you by Evoverse, providing crypto casino source code and case opening solutions, and Wicked Games, creating slot games that slap.