The Czech Republic is building a model for gambling governance

For years debates over gambling regulation have followed a familiar script. Operators complain that politicians legislate by headline. Regulators accuse the industry of resisting oversight. Public-health experts argue that commercial incentives inevitably undermine player protection.  

In Prague, an experiment is under way that attempts to change that dynamic. Recent data suggests that, in the Czech Republic, illegal gambling is costing the market $15.9 million in lost tax revenue, as players have lost up to approximately CZK14.5 billion every year gambling with unlicensed operators.

The Institute for Gambling Regulation (IPRH) is, on paper, an industry association representing around 95% of the Czech Republic’s regulated gambling market. Yet Jan Řehola, the institute’s director, insists it should be understood differently. Rather than acting as another lobbying body, IPRH seeks to function as a permanent platform bringing together operators, addiction specialists, behavioural economists, regulators and government officials to design evidence-based policy. 

Whether this collaborative model ultimately succeeds remains to be seen. But it reflects a broader shift occurring across European gambling regulation: the recognition that neither governments nor operators possess enough information on their own to address increasingly complex problems such as illegal gambling, online harm and cross-border player behaviour. 

The institute’s flagship project, known as IRIS, illustrates that philosophy more clearly than anything else. 

Success means measurable change 

Modern online gamblers rarely confine themselves to a single bookmaker or casino. They maintain multiple accounts, move between platforms and respond to competing promotions. Operators, however, see only their own fragment of that activity. 

“At the core of IRIS is a simple idea,” says Řehola. “We look at gambling from the perspective of the player, not the operator.” 

A customer’s behaviour may appear perfectly ordinary when viewed by one operator in isolation. Across four operators simultaneously, the same customer may be increasing stakes, extending playing sessions and accumulating losses rapidly enough to indicate emerging harm. “It makes no sense for one operator to warn a player while another rewards him with a bonus,” Řehola argues. 

IRIS attempts to solve this collective-action problem without creating what many critics fear: a central database containing personal gambling histories. Instead, the system relies on pseudonymised identifiers assigned by the Czech state licensing infrastructure. Operators never exchange customer names or account details. The platform analyses behavioural indicators across participating operators, classifies risk using a common methodology and returns only the resulting risk category to each operator simultaneously. 

That is the architecture. Whether it delivers is a different question. 

But Řehola is candid about the limits. “I would not present IRIS as a magic solution that works in 100% of cases. Nothing in prevention or addiction risk reduction works like that.” What makes the system different, he argues, is that it allows testing at the market level. “The system continuously evaluates the risk scores of players across all participating operators. We can see whether the overall risk level is rising or falling over time, and whether specific interventions are followed by changes in behaviour.” 

Success, he says, will not be measured by how many messages are sent to players. “Success means measurable changes in players’ behaviour: lower risk scores over time, reduced escalation, better use of limits and fewer players moving into high-risk categories.”

He also acknowledges the flip side: if a certain message or timing does not change behaviour, the model must adapt. “One of the biggest advantages of IRIS is that interventions can be evaluated almost in real time.” 

Are operators really willing to sacrifice revenue? 

That framing raises an obvious question that Řehola does not shy away from: gambling companies are commercial businesses with shareholders, and high-intensity players often generate disproportionate revenue. Why should sceptics believe that commercial incentives have suddenly aligned with player protection? 

“I understand the scepticism,” he says. “It would not be credible to pretend that commercial incentives suddenly disappear because we talk about responsibility.” 

If responsible gambling is left to individual operators acting separately, the incentive backfires: one operator may reduce bonuses and trigger an intervention, while the player simply migrates to a competitor applying lower standards. “In that situation, the responsible operator may lose revenue without actually reducing harm.”

IRIS, he argues, changes that calculation. If the main operators act together under the same methodology, responsible behaviour becomes a shared market standard rather than a competitive disadvantage. 

There is also a longer-term business logic. “Revenue from players who are losing control is not sustainable revenue. It creates regulatory risk, reputational risk and political pressure for much stricter regulation.” From a commercial perspective, retaining players across many years in a pattern that remains sustainable for them may simply be more valuable than short-term extraction that ultimately destroys trust. 

Whether that logic holds consistently under real financial pressure is a question the system has not yet fully answered. 

Is IRIS solving the wrong problem? 

A sharper challenge concerns displacement. If a player responds to IRIS interventions by simply moving to an offshore or crypto casino outside the system altogether, hasn’t IRIS merely shifted the problem rather than solved it? 

“There is a real risk,” says Řehola. “If a responsible-gambling tool in the legal market simply pushes vulnerable players to offshore or crypto casinos, then it has not solved the problem. It has only moved the player into an even more dangerous environment.” His answer is that IRIS cannot be seen in isolation. In the Czech Republic, more than 240,000 people are already registered in the Register of Excluded Persons, and research suggests a significant portion still gamble on the black market.

“People do not stop gambling simply because the state writes their name on a list.” That is why IPRH is working with the Ministry of Finance and the Customs Administration on a national taskforce against illegal gambling. Player protection inside the regulated market and enforcement against the illegal market, he argues, must go hand in hand. 

IRIS is also designed as a proportionate intervention system rather than a punishment mechanism. The objective is not to ban players but to reduce harmful stimuli – aggressive bonuses and targeted marketing – specifically for those in the highest risk category, while leaving ordinary players largely undisturbed. “The aim is to help the player return to a more sustainable level of play while still remaining inside the regulated environment.” 

Who decides what risky gambling actually is? 

A challenge will be the algorithm itself. Behaviour that appears harmful to a psychologist may simply reflect an affluent customer spending within their means. How does IRIS avoid confusing unusual gambling with problematic gambling? 

Řehola is frank. “This question touches one of the biggest limits of any system like IRIS. There is no simple objective truth where a system can say: this person is addicted.” 

His response is that IRIS must not function as a black box making final decisions about people. “It does not diagnose a player and it should not restrict the player’s legal rights. It produces risk signals.” Those signals are then used for proportionate interventions: a reality check, safer-gambling communication, or information about voluntary professional support. 

Crucially, IRIS is not asking only whether someone gambles a lot. “A wealthy player may spend more than an average player and still remain in control. Another player may spend less in absolute terms but show clearer signs of escalation, chasing losses, ignoring limits or losing control over time.” The methodology draws on academic research, international harm markers including the emerging CEN standard, operator data and expert input from addiction specialists and behavioural economists. 

And Řehola acknowledges that perfection is not achievable. “If it flags too many players who are simply unusual but not at risk, we must adjust it. If it misses players who later show clear signs of harm, we must adjust it as well.” Because, in the end, the system’s credibility depends on that honesty about its own limits. 

Cooperation backed by Czech Republic gambling law 

The legal architecture underpinning IRIS is unusual in its own right. Rather than building the technology first and seeking legal approval afterwards, IPRH pursued the opposite sequence. Legislative amendments that entered into force on 1 October 2025 now explicitly permit licensed Czech gambling operators to exchange pseudonymised data for player-protection purposes under carefully defined conditions. 

“The data cannot be used for marketing, commercial profiling or competitive intelligence. It is used for one purpose only: responsible gambling and player protection,”  Řehola explains. 

The legislative change was adopted in parliament, without a single vote against – which is unusual, as gambling policy frequently becomes politically polarised. 

A different regulatory conversation 

Much gambling regulation still alternates between two imperfect instincts. One assumes that commercial incentives inevitably prevent meaningful self-regulation. The other assumes that increasingly restrictive legislation alone can solve problems created by fast evolving digital markets. 

The Czech experiment rejects both. It starts from a simpler idea: good regulation needs information, and that information is spread across governments, operators, doctors and researchers. No one group has the full picture on its own. 

Řehola describes the underlying philosophy as “smart regulation instead of symbolic regulation”. He is careful not to oversell the model’s export potential – each jurisdiction possesses its own legal framework and market structure – but believes the underlying principles travel: cooperation between the legal market, public authorities and independent experts; a strong focus on data; and a commitment to measurable outcomes. 

Whether IRIS ultimately reduces gambling harm remains to be seen, and Řehola is the first to say so. The model must change if the data does not support it. That caution may be its most credible feature.