France’s ANJ algorithm flags 600,000 high-risk gambling players

France’s gambling regulator, the Autorité Nationale des Jeux, has released the first findings from a newly developed algorithm designed to identify potentially excessive gambling behaviour among online and account-based players. The results indicate that harmful gambling patterns are becoming more significant within the regulated market and are increasingly tied to operator revenue.
According to the regulator, approximately 600,000 players in France were identified as showing a high probability of excessive gambling during the second half of 2025. Those players accounted for around €1.2 billion in gross gaming revenue, representing 60% of the total GGR generated within the account-based gambling segment.
The publication forms part of ANJ’s broader 2024-2026 strategic plan, which prioritises the reduction of excessive and pathological gambling across the French market. The regulator stated that the new model is intended to strengthen oversight capabilities while encouraging licensed operators to improve player protection measures.
Regulator highlights growing concentration of gambling revenue
The regulator’s findings suggest that a relatively small share of players is responsible for a substantial portion of market revenue. ANJ stated that the estimated 600,000 high-risk players represented approximately 8.7% of the player accounts covered by the model.
Within that group, around 300,000 individuals were classified as “manifestly excessive” gamblers. ANJ indicated that operators should already be capable of identifying these players through existing monitoring systems and responsible gambling controls.
The regulator expressed concern that the share of revenue generated by high-risk players has continued to rise since 2023. According to ANJ, this trend may indicate that a growing part of the regulated gambling market is increasingly dependent on customers displaying potentially harmful gambling behaviour.
The findings are likely to intensify scrutiny of licensed operators operating within France’s regulated environment. Authorities across Europe have increasingly focused on data-driven monitoring systems as part of broader safer gambling strategies and ANJ’s latest initiative reflects that wider regulatory movement.
Operator identification efforts remain under scrutiny
French gambling operators are legally required to identify problematic gambling behaviour and intervene where necessary. Those interventions can include direct communication with customers, the introduction of personalised deposit or betting limits, referrals to specialist support organisations and account suspensions or closures in severe cases.
ANJ acknowledged that operators have improved their detection efforts over the past year. According to the regulator, the number of players identified by operators as excessive gamblers rose from approximately 31,000 in 2024 to around 89,000 in 2025.
Despite that increase, the regulator stated that current identification levels remain below expectations when compared with the estimates generated by its own algorithmic model.
The gap between operator-reported figures and the regulator’s estimates is expected to become an important issue in future compliance reviews. ANJ indicated that operators will likely face increasing pressure to demonstrate that their monitoring systems are capable of detecting risky gambling behaviour more effectively.
How the ANJ algorithm was developed
The algorithm was created using player-account data continuously transmitted to the regulator by licensed operators, as well as by France’s state-linked gambling groups, FDJ UNITED and PMU.
ANJ stated that work on the project began in 2024 and incorporated findings from scientific research focused on gambling-related harm and behavioural risk indicators.
The model evaluates players using 23 separate indicators before assigning each individual a risk score. Those indicators include gambling frequency, financial transaction patterns, use of gambling control tools and historical behavioural data linked to player activity.
Players are then placed into four categories:
- Recreational
- Moderate risk
- Excessive
- Manifestly excessive
According to the regulator, the model’s accuracy was assessed using the Canadian Problem Gambling Index. ANJ also noted that the project was reviewed under the supervision of a scientific committee.
The regulator described the tool as one of the first large-scale systems of its kind to be introduced within Europe’s regulated gambling market.
ANJ plans to use the model as a compliance benchmark
ANJ stated that operators will be permitted to use the algorithm voluntarily alongside their own internal monitoring systems. The regulator believes the model can serve as a benchmark that allows operators to compare their detection capabilities against a regulatory reference framework.
The regulator also confirmed that the tool will play a role in future compliance assessments. During the next review cycle in 2027, ANJ may compare the number of excessive gamblers identified by operators with the figures generated through the regulator’s algorithm.
This development could significantly affect how operators approach safer gambling obligations in France. Businesses may face increased expectations regarding customer monitoring technologies, intervention strategies and evidence-based compliance reporting.
Industry observers believe that data-led enforcement tools are likely to become more common across European gambling markets as regulators seek measurable evidence that responsible gambling systems are functioning effectively.
Retail gambling sector may face similar standards
Although the current model focuses on online and account-based gambling activity, ANJ also indicated that similar standards should eventually apply to retail gambling environments.
This could affect land-based networks operated by monopoly-linked groups and physical points of sale throughout France.
ANJ President Isabelle Falque-Pierrotin commented on the launch of the new system, stating:
“The finalization of this algorithm and its release to operators represent a decisive milestone for the regulator. It demonstrates its capacity to design an innovative and high-performing tool, tailored as closely as possible to the reality of online players’ habits.
Alongside barometric studies, the algorithm provides objective data on the efforts to identify excessive players – efforts that operators must pursue without delay. Finally, it is essential that this identification can also be implemented at physical points of sale, an objective we have been asking the two monopolies to pursue since 2024.”
Her remarks underline ANJ’s intention to expand supervisory expectations beyond online gambling alone. The regulator appears determined to establish a more comprehensive framework capable of monitoring gambling-related harm across multiple channels.
European regulators increasingly focus on data-led oversight
France’s latest initiative reflects a wider trend among European regulators toward greater use of technology and behavioural analytics in gambling oversight.
Several jurisdictions, including Spain and the Netherlands, have explored similar approaches aimed at identifying high-risk gambling behaviour earlier and improving intervention systems. However, ANJ stated that its model is among the first to be formally introduced as an operational regulatory benchmark within Europe.
The growing use of algorithmic supervision is likely to raise important discussions around privacy, proportionality and regulatory accountability. At the same time, regulators argue that increasingly sophisticated monitoring systems are necessary to address evolving gambling risks within digital markets.
For operators, the introduction of advanced regulatory tools may create additional compliance obligations and operational costs. Companies could be required to invest further in artificial intelligence systems, player behaviour analytics and internal responsible gambling teams.
Wider implications for the French gambling market
The publication of the first results from ANJ’s algorithm marks an important development for the French gambling sector. The regulator has made clear that safer gambling measures will remain central to its regulatory strategy over the coming years.
The findings also highlight the financial importance of high-risk players within the regulated market. With 60% of account-based gross gaming revenue linked to players flagged as potentially excessive gamblers, the issue is likely to remain a key area of political and regulatory attention.
For gambling operators, the message from ANJ is increasingly direct. Regulatory authorities expect stronger identification systems, faster interventions and measurable outcomes linked to player protection policies.
As European regulators continue to strengthen oversight frameworks, France’s approach may become an influential example for other jurisdictions seeking more data-driven methods of monitoring gambling-related harm. The success or limitations of ANJ’s algorithm will therefore be closely watched by policymakers, operators and responsible gambling experts across the wider industry.
Conclusion
France’s ANJ has taken a significant step toward expanding regulatory oversight through the introduction of its new gambling risk algorithm. By identifying approximately 600,000 players as potentially excessive gamblers, the regulator has highlighted both the scale of gambling-related harm concerns and the growing dependence of market revenue on high-risk customer activity.
The initiative also signals a broader shift toward data-led supervision within regulated gambling markets. Operators are now facing increasing expectations to improve monitoring systems, strengthen intervention processes and provide clearer evidence that responsible gambling measures are working effectively.
While the long-term impact of the model remains to be seen, the release of these findings places additional pressure on the French gambling industry to prioritise player protection and regulatory compliance. It also reinforces France’s position as one of the more proactive European jurisdictions in the development of technology-based gambling oversight.
FAQs
What is ANJ in France?
ANJ is France’s national gambling regulator responsible for supervising licensed gambling operators and enforcing responsible gambling rules.
How many players were flagged by the ANJ algorithm?
The regulator stated that around 600,000 account-based players were identified as potentially excessive gamblers during the second half of 2025.
What percentage of gambling revenue came from high-risk players?
According to ANJ, the identified high-risk players generated approximately 60% of gross gaming revenue within the monitored segment.
What does the new ANJ algorithm analyse?
The model evaluates 23 risk indicators including gambling frequency, financial activity, player history and use of responsible gambling tools.
What categories does the algorithm use?
Players are classified into four groups: recreational, moderate risk, excessive and manifestly excessive.
Are French gambling operators required to identify harmful gambling behaviour?
Yes. French law requires licensed operators to identify problematic gambling behaviour and implement appropriate intervention measures.
How many excessive gamblers did operators identify independently?
Operators reportedly identified around 89,000 excessive gamblers in 2025 compared with 31,000 in 2024.
Will the algorithm become mandatory for operators?
ANJ stated that operators can use the tool voluntarily alongside their own internal monitoring systems.
Could retail gambling locations also face similar monitoring standards?
Yes. ANJ indicated that similar identification expectations may eventually apply to physical gambling points of sale.
Why is the ANJ algorithm important for the European gambling industry?
The system represents one of the first large-scale regulatory tools in Europe designed to measure gambling-related harm using behavioural data analytics.
Related Posts

Will Mace joins Mindway AI to advance player safety
June 25, 2026







































