EN 18144 and the challenge of turning markers of harm into action
- 20 hours ago
- 4 min read
EN 18144 has now been formally published. The voluntary European standard sets out nine common behavioural markers of harm in online gambling, giving the sector a shared point of reference across markets that have previously taken different approaches.
Publication does not make EN 18144 a legal requirement or a condition of national gambling licences. It was developed through the European Committee for Standardisation (CEN) and is available through national standards bodies.
The standard may influence governance, procurement and assurance. It gives boards a basis for reviewing monitoring systems and operators a way to examine supplier claims. Regulators may also consider it when assessing how operators identify potentially harmful gambling behaviour.
Over time, EN 18144 could become a benchmark for assessing markers-of-harm systems. Operators taking a different approach may be asked to show how their systems address the same areas of risk.
EN 18144 followed a 2022 proposal from the European Gaming and Betting Association (EGBA), an industry body representing online gambling operators in Europe. Operators, national authorities, researchers and harm-prevention stakeholders contributed to its development. Its nine markers cover changes in stake volume or frequency, intensity of play, deposits, withdrawals, player-initiated contact, session behaviour, multiple-product use, losses and changes to safer gambling controls.
Many operators already monitor deposit increases, longer sessions, cancelled withdrawals and rising losses. EN 18144 brings those indicators together in a common European standard.
Terminology, regulatory expectations and approaches to customer interaction vary between European markets. EN 18144 provides a common foundation that can be used alongside local regulatory and operational requirements.
EN 18144 may also improve conversations between operators and their technology providers. References to artificial intelligence, real-time monitoring and proprietary risk models do not always explain how a system uses behavioural data. Operators can use the standard to ask which markers are monitored and what happens when they are detected.
The answer cannot end with a checklist.
A supplier may show that every marker is captured somewhere across its monitoring, customer-service and case-management systems. Technical coverage alone says little about whether the information is brought together, weighted consistently or escalated in time for action.
Collecting the data is one part of the process. The harder task is interpreting what it means.
A sudden increase in deposits may indicate a material change in behaviour or reflect a temporary change in circumstances. A long gambling session may warrant attention, depending on the customer, the product, the amount spent and what else happened during the session. Gambling across several products may point to increasing intensity or ordinary use of an operator’s website.
An individual marker may be enough to start a review. The decision that follows should consider the other information available.
A monitoring system should assess changes over time, consider the customer’s established activity and recognise when several markers appear together. A £500 deposit cannot be interpreted in isolation. Its relevance will depend on whether it differs from the customer’s previous activity and what other indicators are present.
EN 18144 does not prescribe a universal threshold for every operator, product and market. A single European trigger could generate large volumes of alerts while overlooking less obvious changes in individual behaviour.
Operators still decide how the markers are applied. Two businesses can monitor the same nine markers and set very different thresholds. A system may prompt an early review or wait until the behaviour becomes extreme. Coverage of the same markers can therefore result in different levels of protection.
Mapping a system against the nine markers can show coverage without explaining how well that system works in practice.
Operators should look beyond claims that a product covers all nine markers. They need to understand the data used, the periods over which behaviour is compared, how indicators are weighted and when an alert is escalated.
Any model has limits, and some produce more alerts than teams can reasonably review. However, others set the bar so high that changes in behaviour never reach a person. What matters is that the right cases are picked up in time.
Identifying a customer for review should trigger a defined process. The case needs an owner, an assessment of the available information and a decision on whether any intervention is required.
Any response should remain proportionate. An individual marker may lead to information, a reminder of available controls or a carefully framed conversation with the customer. Where concerns continue or become more serious, the operator may consider limits, further enquiries, suspension or closure.
The process should also record what happened after an intervention. This includes whether the customer’s behaviour changed, limits were put in place, the concerning pattern continued, or further action was needed. Recording an interaction alone does not show whether it worked.
Automated systems can process large amounts of data, identify changes and help teams prioritise cases. They cannot replace judgement. Someone still has to interpret the information, consider the customer’s circumstances and decide what should happen.
Boards need more than headline totals showing how many alerts or interactions occurred. High numbers may reflect sensitive thresholds, increased activity or more customers showing patterns of concern. Without context, the figures say little about how well the system is working.
They should understand why customers were identified for review and how quickly cases were assessed. They also need to know which interventions were used, whether alerts recurred and what happened afterwards. Outcomes should be compared across customer groups, products and markets.
A supplier may provide the monitoring technology, but the licensed operator remains responsible for its use. Claims of EN 18144 coverage need to be supported by an explanation of how the model works, how alerts are produced and how decisions are made.
The publication of EN 18144 is a genuine achievement. It gives the European gambling industry a common vocabulary and baseline. It can make monitoring systems easier to compare, help operators examine supplier claims and help regulators see where important indicators are being missed.
Markers of harm can draw attention to changes in gambling behaviour that warrant closer review. EN 18144 gives the industry greater consistency in what it monitors. The quality of protection will depend on how operators interpret the markers, decide when to intervene and review what happened afterwards.
Earlier in this series:
