Spain’s DGOJ opens consultation on AI monitoring project with no guarantees on the timeline or testing of an ambitious gambling harms detection system. 

DGOJ, the Directorate General of Gambling in Spain, has announced that it will include 60 specific behavioural and transactional variables in its AI-powered monitoring project, which aims to analyse gambling patterns and identify the triggers behind high-risk or problematic play. The system represents a landmark effort by the Spanish regulator to introduce real-time behavioural oversight across the country’s licensed online gambling operators.

The adoption of this AI-based mechanism was established as a central pillar of the Royal Decree for Safer Gambling Environments (2023), a legislative framework that tasked the DGOJ with designing technological interventions to prevent gambling harm and protect vulnerable users. Backed by Spain’s Ministry of Consumer Affairs, the project sets out to implement an AI-driven system capable of identifying early signs of risk and enabling targeted, one-to-one operator interventions.

Back in 2023, the DGOJ initially framed the initiative as a model for Europe, with the Spanish government claiming that the country would soon have the “most sophisticated gambling monitoring system of any EU member state.”

However, stakeholders voiced concerns over the project’s complexity and lack of transparency, particularly as no public information was made available throughout 2024. These concerns were partially addressed during the July 2025 Gaming in Spain Conference in Madrid, where DGOJ President Mikel Arana reassured delegates that development was ongoing.

“This remains a regulatory priority. We are building the most robust technological foundation possible — one that can serve the market long-term,” Arana said.

Public consultation on risk variables

This week, the DGOJ published its formal public consultation on the technical design of the system, inviting contributions from operators, tech providers, academics, and the public. Open until 25 September 2025, the consultation is aimed at refining the system’s parameters and assessing its technical and legal feasibility.

The model, which will be built using XGBoost (Extreme Gradient Boosting) machine learning, is set to analyse more than 60 data variables. These include the frequency and intensity of gambling sessions, changes in stakes after wins or losses, and the volume and timing of deposits and withdrawals. 

The DGOJ also plans to monitor cash-out frequency, duration of play over consecutive days, and betting preferences, specifically among players engaged in live, or in-play, wagering.

Spanish authorities  demand that monitoring focus on risks posed by in-play betting, due to its rapid pace and the emotionally reactive decisions it encourages. By synthesising behavioural indicators, the AI model will assign player-specific risk scores. These, in turn, will trigger alerts and enable uniform, data-led interventions by operator support teams.

Timeline uncertainty

Despite its ambition, the project’s current design has raised questions over scope and implementation timelines. The AI mechanism will only apply to the private online gambling market—covering verticals such as sports betting, casino games, poker, roulette, blackjack, and bingo — but will exclude lotteries, even though lottery products remain the most played form of gambling among Spanish adults.

This means that state monopolies like SELAE and ONCE are not subject to the same AI monitoring requirements, an omission that has drawn sharp criticism from industry commentators.

“Excluding lotteries undermines the credibility of a system that claims to monitor public health risks,” noted SBC Noticias in a recent editorial. “For a country where lottery is the most common form of gambling, it’s a glaring regulatory blind spot.”

Adding to the concern, the DGOJ has not provided a clear timeline for the AI system’s rollout. While the Royal Decree gives a two-year window for development, suggesting a potential launch by late 2025 — there is no confirmed date for beta testing, nor has the regulator clarified how risk scoring will be standardised across platforms.

Implementation may also be affected by Spain’s next general election, which must be held no later than summer 2027, adding further political uncertainty to the regulatory landscape.

Regulatory gaps persist

The AI monitoring initiative is being developed in parallel with unresolved legal challenges facing the DGOJ. In particular, Spain’s High Court has struck down several key provisions of the country’s gambling advertising code, including restrictions on celebrity endorsements and social media promotions — a decision that has yet to be addressed with an updated code of conduct.

“Without clarity on advertising and enforcement timelines, the AI project risks being another well-intentioned initiative delayed by bureaucracy,” warned SBC Noticias, calling for a more holistic approach to gambling harm prevention.

Despite these outstanding issues, the AI system reflects a serious shift toward technology-led regulatory oversight in Spain’s gambling market. If successfully implemented, it could serve as a blueprint for other EU jurisdictions exploring algorithmic models of risk detection. For now, however, the project remains in development — with its full effectiveness and independence yet to be tested.

__

September 15 will see SBC organise a ground breaking charity football event in Lisbon. Make sure you get the chance to see some of the most legendary names in football by securing your ticket today at https://www.legendscharitygame.com/