Free-To-All Harm Detection Algorithm Promises To Shake Up Player Protection

A Dutch university, with involvement from Spain’s regulator, has released a free algorithm which promises to detect risky gambling and allow operators to better protect consumers, in an open-source challenge to the privately owned responsible gambling industry.

The University of Amsterdam (UvA) has created an algorithm that tracks player behaviour and uses it to report on how risky their behavior has become.

It was developed by using data and funds provided by the Netherlands Gambling Authority (KSA), taking in online casino wagers from 13 different providers across two years.

Researchers at UvA drew on information collected between July 2023 to July 2025.

Licensed operators in the Netherlands are required to link up to a “data vault” managed by the KSA in order to send pseudo-anonymised wagering records to the regulator.

The data is used to monitor for potential compliance breaches, but can also be used for research. This project was the first such use of the vault by an academic institution.

A team led by PhD candidate Charles de Leau at UvA harvested 24 months of this information to train their algorithm.

"The fact that we were able to analyze all bets from 13 different casinos over two years has never been done before by independent researchers," said De Leau.

"With this massive amount of data, which is normally used by casinos themselves for marketing purposes, we can see for the first time on such a scale which patterns in gambling behavior often precede serious problems," he said.

The academics note that until now, virtually all similar systems had been offered commercially on the private market and were developed in partnership with gambling companies.

“With a public and independent model, regulators worldwide now have their own transparent frame of reference at their disposal, free from commercial interests,” UvA said.

Spanish influence

The Dutch project comes at the same moment as Spain’s regulator, the DGOJ, has been working on its own harm detection algorithm.

Once it is completed, it will become mandatory for all licence-holders in the Spanish market to adopt the DGOJ’s system, or a similar version which they can demonstrate reaches the same standards of player protection.

The KSA said that UvA’s project drew “much inspiration” from the DGOJ and the university notes that its algorithm was developed in “close collaboration” with Spanish officials.

This suggests that the algorithm that is eventually rolled out by the DGOJ will bear significant similarities to this Dutch project.

UvA is also encouraging more regulators internationally to adopt a version of its system.

“The algorithm is open source: the code and methodology are public. This means that other researchers can verify its operation and build upon it, and that regulators in other countries can also start using the model as a reference for supervising online casinos.” the university said.

Duty of care supercharged

In the Netherlands, Dutch gambling providers owe a legal duty of care to their customers.

Failures by operators to live up to that standard have resulted in millions of euros of fines so far.

As recently as June this year, the KSA issued a fine of €886,000 to 711 B.V. for duty of care breaches, based on analysis of the gambling patterns of ten high-spending individuals.

The UvA notes that how operators choose to handle high volume gamblers is the most critical factor in determining compliance with their duty of care requirements

It pointed to research by French authorities that found that a small group of higher-spending players are responsible for 60 percent of revenues.

Although few nations outside the Netherlands impose a legal duty of care on their licensed operators even in markets like the UK — where a recent court case established that a duty of care definitely does not exist — operators can still expect to be sanctioned for failing to monitor and protect their customers.

The algorithm assigns a risk score to players based on their gambling behavior.

Although the UvA revealed that it was trained with online casino data, the KSA’s website states it can be deployed for all forms of gambling.

The regulator is also expected to use the algorithm itself to test the data it collects in its vault.

It can then, during compliance assessments, ask to see an operator’s risk scoring and player protection timeline to see if that matches up with how investigators believe the licensee should have handled the gambler.

'This is a very different way of looking at player protection, and that will be quite a shift for many parties,” said De Leau.

“In my opinion, however, it is incredibly important in this rapidly growing market that we are given more opportunities to protect players.'

Any interested parties, including Dutch licence holders, can access the algorithm on the KSA’s website.