Recently, experts from Totalizator Sportowy, the Polish Bank Association and the National Clearing House called for a new approach to addressing illegal online gambling. They suggest that the next phase of enforcement should emphasise financial data analysis, the use of artificial intelligence, and enhanced cross-sector collaboration.
A debate was held on cooperation between Totalizator Sportowy and financial institutions. Artificial intelligence was a key topic. Wojciech Pantkowski, Vice President of the National Clearing House, noted that the growth of instant payments has greatly reduced the time available to conduct effective risk analysis.
He explained that ensuring transaction security requires verifying the sender, recipient, and beneficial owner, but in today’s instant transfer systems, this process must be completed in as little as 10 seconds.
Separately, it was reported that the Polish Sejm, the lower house of parliament, passed a law criminalising "patostreaming" - the online distribution of illegal content, including gambling-related content. However, this proved to be misleading, as the version of the bill ultimately passed did not include a separate criminal offence for streaming gambling-related content.
Global Gaming Insider reached out to Marek Plota, online gaming legal expert, owner of RM Legal Law Firm and founder of Gaming in Poland, for comment.
What are the main barriers to closer cooperation between banks, regulators and gaming authorities in using financial data analysis and AI to combat illegal gambling?
The main barrier is not the absence of legal tools, but the fact that the stakeholders often have different interests. The regulator, Totalizator Sportowy (which holds the state monopoly in the online casino sector) and licensed bookmakers all share a common objective which is reducing illegal online gambling.
For banks and payment institutions, however, the picture is more complex. Enhanced cooperation typically means additional legal obligations, compliance responsibilities and significant technological investment in developing monitoring, screening and transaction-analysis tools. It may also inevitably result in the loss of transaction-related revenues generated by customers engaging with gambling operators. While these interests are not necessarily contradictory, they do create different priorities and levels of urgency among the stakeholders involved, but they are not always perfectly aligned.
At the same time, we should acknowledge the progress already made. Cooperation between the Polish Financial Supervision Authority (KNF), Totalizator Sportowy and representatives of the licensed betting sector has already led to a visible regulatory reaction, including warnings addressed to the payment market. We also see broader consultations with the Polish Bank Association and payment service providers aimed at identifying and blocking transfers connected with illegal gambling. These may be small steps, but in this area every step matters.
How feasible is it in practice to integrate banking data into gambling enforcement efforts without breaching privacy or data protection rules?
In my view, the privacy and data protection risks are fully manageable. Payment institutions have legitimate grounds to detect and respond to potentially unlawful activity. Those grounds arise from AML rules, gambling law, criminal law and their broader compliance obligations.
The key point is that we are not speaking about unrestricted surveillance of customers. We are speaking about risk-based monitoring of transactions that may indicate support for illegal gambling activity.
Where red flags appear, institutions should be able to pause or reject transactions, increase the risk level, perform enhanced checks and notify the competent authorities.
Cooperation between the Polish Financial Supervision Authority (KNF), Totalizator Sportowy and representatives of the licensed betting sector has already led to a visible regulatory reaction
How realistic is the use of AI and machine learning in identifying illegal gambling transactions in real time?
The use of AI for identifying potentially illegal gambling transactions is entirely realistic and, in my view, does not face any fundamental legal obstacles, provided that the system is used as a compliance-support tool and not as an uncontrolled automated enforcement mechanism.
The AI Act does not prohibit systems used to support a human assessment of possible involvement in criminal activity where that assessment is based on objective and verifiable facts linked to that activity.
In practical terms, AI could be useful in detecting patterns that are difficult to identify manually, such as recurring payment chains, suspicious merchant structures, rapid changes of payment intermediaries, links with domains from the prohibited register or discrepancies between the declared business activity and actual transaction flows. The output should then be reviewed through a controlled compliance process.
Are financial institutions in Poland technologically prepared to flag gambling-related risks within a 10-second transaction window?
It would, of course, be ideal if transfers to illegal casinos were detected before execution. However, the 10-second transaction window should not be treated as the only meaningful benchmark. Even if full real-time blocking is not always technically feasible, retrospective analysis is still highly valuable.
A very useful enforcement model would be to identify suspicious flows after the transaction, trigger KYC or enhanced due diligence on both sides of the payment chain, establish the real identity of the recipient and escalate the case where necessary. Regular explanatory proceedings and repeated interventions by payment institutions and public authorities would reduce conversion to illegal casinos.
Once customers and intermediaries see that such transactions are being questioned, the appetite for participating in illegal gambling decreases.
How quickly can illegal operators realistically adjust to financial restrictions compared to domain blocking?
Illegal operators can adjust to domain blocking almost immediately. In many cases, it requires little more than cloning the domain and directing users to a new address. Sometimes, even active communication with users is not necessary because the ecosystem is already prepared for frequent domain changes.
Financial restrictions are more difficult to bypass, because they affect the operator’s ability to convert traffic into deposits and withdrawals. They will not eliminate the grey market overnight, but they can create friction, increase operational costs and reduce conversion.
The same is true of the prohibited domain register. Each individual measure may have only a limited effect, but the cumulative impact matters. In this area, progress is slow, but consistent pressure can gradually reduce the scale of the problem.
In practical terms, AI could be useful in detecting patterns that are difficult to identify manually, such as recurring payment chains, suspicious merchant structures and rapid changes of payment intermediaries
Regarding the “patostreaming” legislation, why was the provision covering illegal gambling content removed?
In the version of the bill that has been passed by the Sejm and forwarded to the Marshal of the Senate and the President, the provision no longer introduces a separate criminal offence for streaming gambling-related content.
However, the removal of gambling from this amendment should not be read as a policy decision to tolerate the phenomenon. It is a formal correction. Polish law already contains fiscal-criminal provisions penalising illegal advertising and promotion of gambling, in particular under the Fiscal Penal Code.
Adding the same conduct to the Criminal Code through the patostreaming amendment would have duplicated the existing regime and created serious problems of proportionality and legal certainty.
Which existing legal tools and enforcement mechanisms does Poland currently use to combat the online promotion of illegal gambling?
The fact that gambling was removed from the patostreaming bill does not mean that the state is withdrawing from enforcement.
It means that the more appropriate legal route is to use, and where necessary improve, the existing gambling and fiscal-criminal framework rather than to duplicate offences in the Criminal Code.
In the gambling sector, Poland already relies on a separate enforcement architecture, including licensing requirements, the register of prohibited gambling domains, restrictions on payments to illegal operators, AML obligations, and fiscal-criminal liability for unlawful advertising or promotion.
Estimates suggest that 20–30% of the Polish online casino segment remains unregulated
