Problem gambler AI detection
The gambling industry is increasingly using artificial intelligence to target prospective punters and predict customer habits and maxmise revenue and profit, a move that is seen as controversial in some quarters.
At the same time, machine learning and related technologies are being used to detect problem gamers. However, it remains unclear how effective these solutions are.
UK Anonymous Player Awareness System
In November 2019, the UK Betting and Gaming Council, an industry group representing 90% of the UK betting and gaming market, launched an AI-powered Anonymous Player Awareness System (APAS) designed to detect and prevent problematic behaviour in players. The system locks gamblers out of machines for 30 seconds if erratic or excessive play is detected.
Experts said the APAS was a move in the right direction, but were concerned that 30 seconds is long enough to have any real impact. A 2019 study of Norwegian gambling machines found that the break caused 'no significant effect' on the amount of money staked during a subsequent gambling session or how long that session lasted.
Betfair risk assessment algorithm
In June 2021, Luke Ashton, from Leicester, UK, committed suicide after racking up large debts had been categorised as a 'low-risk' customer by a Betfair algorithm that had 'found nothing in his betting patterns that would trigger human intervention that might have restricted his gambling.' The coroner ruled Betfair had failed to meaningfully interact or intervene when Mr Ashton's gambling activity spiked.
Gambling venues across New South Wales, Australia, have introduced facial recognition to detect people who had voluntarily signed up to a 'self-exclusion' scheme for problem gamblers. Digital rights activists complained the technology is 'invasive, dangerous and undermines our most basic and fundamental rights'.
Operator: Flutter UKI/Betfair; Ladbrokes/Coral
Developer: Betting and Gaming Council (BGC); Mindway AI; Optimove
Country: Australia; UK; USA
Purpose: Detect problem gamblers
Technology: Machine learning
Issue: Accuracy/reliability; Effectiveness/value; Privacy; Safety; Surveillance
Transparency: Governance; Black box; Marketing
UK Gambling Related Harm All Party Parliamentary Group (2020). Online Gambling Harm Inquiry - Final Report (pdf)
Spencer Murch W., Kairouz S., Dauphinais S., Picard E., Costes J-M., French M. (2023). Using machine learning to retrospectively predict self-reported gambling problems in Quebec
News, commentary, analysis
Published: June 2023