In the culmintating part of our latest roundtable, the going challenges associated with an increased pace of AI implemenetation is detailed, in addition to an elaboration on how entertainment and personalisation will surely be enhanced as a result.
Dr. Andreas Koeberl, Chief Executive Officer at BetGames, Oleksii Mylotskyi, Chief Technology Officer at Playson, Thomas Smallwood, Chief Commercial Officer at OpenSlots, and Bjørnar Heggernes, Chief Commercial Officer at The Mill Adventure, once again offer their insights.
Q3. What are the key challenges in integrating AI tools across the value chain? From data quality and infrastructure to cost and player trust.
Dr. Andreas Koeberl: Everything in AI depends on data: the data used to train models, the data being ingested, and so on. That’s the most important component. Market readiness is a major issue as well. Most users aren’t prepared to take full advantage of it yet.
Infrastructure will eventually become a commodity, so that won’t be the challenge. Player trust will sort itself out over time, especially as users start leveraging AI tools themselves, potentially quicker than most studios and operators react.
Oleksii Mylotskyi: Integrating AI across the iGaming value chain comes with several major challenges. First, data access – as a supplier, Playson has no player data, this protects privacy but limits deep personalisation.
Second, infrastructure and cost – AI requires significant processing power, so scaling sustainably means balancing innovation with efficiency. Finally, player trust – AI must enhance, not manipulate, experiences. Transparency about its use is critical to maintaining player confidence and long-term engagement.
Thomas Smallwood: As with any resource, the key is to use AI where it can be genuinely helpful and not for the sake of it. It is not a magic wand but an enabler.
OpenSlots, for example, uses AI to help reduce the length of development roadmaps drastically and enable instant replication and cheap prototyping, which can be gamechangers for game creators.
The challenge is knowing where AI adds genuine value versus where traditional methods still work best.
Bjørnar Heggernes: The approach needs to be precise and calculated based on informative decisions. This usually comes down to if the good outweighs the bad. How does the tool solve current problems, is it reliable and accurate, and would it financially be the right solution?
One major challenge is data quality and consistency. AI models are only as strong as the data they process, so if the input is noisy, outdated, or poor, the output will be too.
Imagine connecting an AI assistant to a compliance Knowledge base that lacks key details, updated two years ago or just suffers from inconsistencies – even with AI, the documentation needs to be maintained, as AI does not suddenly just replace the people, but rather helping the person on the other side extracting the key information they need. Integrating AI across multiple systems and teams requires alignment on data formats, workflows and compliance.
Cost and scalability are also constant considerations. If AI tools become too expensive to build and maintain in-house, the smarter choice may be to integrate with specialists who already have strong data layers and optimised pipelines in place. This approach can improve data quality, streamline costs and apply more selective, relevant data while reducing noise.
In addition to that, governance and data privacy are two aspects that are very important to be taken in consideration when planning adoption and roll-out throughout any organisation.
Q4. How, as an industry, do we ensure AI is enhancing entertainment and security, rather than creating unfair advantages?
OM: On the entertainment side, AI empowers studios like Playson to create deeper, more immersive experiences that adapt to player preferences while maintaining fairness. From the security perspective, AI strengthens responsible play, detects potential fraud and supports AML processes by identifying abnormal patterns early.
To maintain trust, independent certification of AI systems, just like RNGs today, will be essential, giving regulators, operators and players confidence in every interaction.
TS: As mentioned above, we use AI where it can make a difference to deliver a better service and product, not to cut corners. We work in a highly regulated industry, which we welcome, and don’t see any conflict in game development between regulation and the use of AI to deliver higher quality, faster service to our customers.
BH: I touched upon some pointers previously relating to curating content with the help of AI, by tapping into data points to understand player preferences better. I do not see the use of AI creating unfair advantages for anyone – was the calculator unfair when that was introduced? Of course not. The challenges came when different institutions had different rules and failed to understand how to implement it effectively.
It is the same here. The responsibility lies in how we, as an industry, apply AI. If used well, it enhances entertainment by making content more relevant and engaging, while strengthening security through better player identification and protection systems.
The goal should always be to give players the best possible experience in a safe environment. This is our collective responsibility – operators, suppliers, regulators and marketeers need to align on transparent and ethical use of AI. That is how we ensure it serves everyone equally, not unfairly.
AK: Fraud will always be a major issue, and AI will not eliminate it. In fact, it will likely make it worse. The only way to manage this is by training people to use AI properly. Better yet, require them to use it. In more aggressive words: any engineer unwilling to adopt it should leave the organisation.
That sounds controversial, but what we need to appreciate this new pace. This is not a technology that leaves much room for reaction after the fact. Market readiness in our industry is low, mainly due to limited talent density, so the sooner we start, the better.
Q5. Looking forward, how might partnerships between slot studios and tech providers evolve to unlock the full potential of AI personalisation?
TS: There are many stages to producing a slot, and I believe some specialists will concentrate on one aspect of it, such as mathematics and mechanics, to potentially deliver very specific tools to the industry. We will see an ecosystem of tech providers offering different elements of the wider process, creating more specialised and efficient solutions for everyone involved.
BH: Through data. This is considered the most valuable asset in the world right now, even with BTC and gold reaching new heights; data has an indefinite value.
The challenge is that many limitations exist around what data is collected, shared and processed. Some are due to cost restraints, others due to privacy laws, technical barriers, or simply the lack of structured collection. For AI personalisation to reach its full potential, slot studios and tech providers need to collaborate more openly around data, and responsibly, of course.
If studios shared more relevant behavioural data with platforms and third-party tools, even basic AI models could deliver far more accurate personalisation. In return, tech providers can share aggregated demographic and performance insights that do not compromise player privacy. This kind of two-way data collaboration removes the guesswork behind why some games thrive while others fade.
Trends like Starburst’s longevity, the rise of Megaways, or the popularity of high-volatility slots and bold themes set by Nolimit City, all tell us that player preferences evolve. Rather than chasing what the masses might like next, the focus should be on using data – enhanced by AI – to understand individual preferences and deliver experiences that treat players as people, not just numbers in a spreadsheet.
AK: The obvious one here is vibe coding. This will likely happen within the next six to 24 months. Studios will start offering platforms that support on-demand reskinning, and possibly even full slot creation, from natural language prompts.
I’ve seen a few early concepts, and while they’re still rough at the moment, the pace of GenAI’s progress makes the timeline above feel realistic. I strongly believe this will spark new operating models. It inevitably has to.
OM: Partnerships between slot studios and tech providers should focus on shared innovation and responsible growth.
For operators, the supplier’s goal should be delivering AI-ready games and promotional tools that integrate seamlessly into their ecosystems, offering both adaptability and compliance.
We have seen the first test products on the market, but it is still not widespread yet. With tech providers, a selective approach should be taken, adopting AI infrastructure that amplifies creativity and regulatory precision.
And with regulators, open collaboration should be established to ensure that AI-driven innovations remain transparent, fair and always aligned with player protection – unlocking the full, ethical potential of personalisation in iGaming.
