for early access to the latest igaming videos!
and get the latest igaming news first!
The regulator says some gambling companies are relying on algorithms they don’t fully understand.
A growing concern in operator compliance
Artificial intelligence is transforming compliance, but not always for the better.
The British Gambling Commission (BGC) has issued a warning to operators after discovering widespread weaknesses in how AI-powered systems are being used to manage anti-money laundering (AML) controls.
According to its latest compliance report, the regulator found that many operators have implemented AI, algorithms, and behavioural models to flag potential money-laundering and terrorist-financing (MLTF) risks, but some cannot clearly explain how those systems actually work.
The tools, often used to produce aggregated “risk scores” for customers, are designed to detect unusual behaviour. However, the Commission said several operators had failed to properly configure them or test their accuracy before integrating them into live AML systems.
When automation goes wrong
Some customers were able to move large sums before systems even noticed.
Investigators found cases where high-risk indicators were missed entirely because of algorithm configuration errors. In other instances, models required a minimum level of account activity before they would trigger a risk score, leaving a blind spot in the early stages of a customer relationship.
The Commission warned that this oversight can allow users to transact large amounts of money before controls take effect.
It also discovered that in some firms, compliance teams could not see why an algorithm had flagged a customer, making it impossible to address the risk effectively.
“When implementing AML controls, operators need to ensure their systems reflect the risks identified in their MLTF assessments,” the Commission said. “Controls must be appropriate, transparent, and implemented effectively.”
Outsourcing and oversight issues
Some operators are still handing too much power to third-party consultants.
The report also revealed that several gambling companies had outsourced their risk assessments and control documentation to external AI and consultancy providers, sometimes resulting in incorrect or incomplete information being used to guide compliance procedures.
In response, the BGC confirmed it will review the effectiveness of operators’ AI tools as part of upcoming assessments. It said teams using these systems must have access to methodology details, including what the algorithm measures, how scores are weighted, and why certain customers are flagged.
“During the customer review part of the compliance assessment, we will examine how the algorithm operates in practice in relation to specific customers, and test whether this is appropriate and effective.”
Beyond AI: persistent AML weaknesses
Technology isn’t the only compliance problem the Commission is watching.
Alongside its AI concerns, the regulator pointed to recurring failures in AML risk assessments. Some operators had not considered all relevant risks in line with the Commission’s guidance, while others showed a disconnect between written policies and day-to-day controls.
The report also found examples where customer risk profiles were incomplete or based solely on spend thresholds, ignoring other high-risk indicators such as geography, occupation, or product type.
“When implementing AML controls, operators need to ensure that their controls address the risks identified in their MLTF risk assessment, and that such controls are appropriate to the business and implemented effectively.”
Documentation and record-keeping standards also came under scrutiny. The Commission cited poor oversight from Personal Management Licence (PML) holders and inconsistent reviews of customer documentation, including unverified third-party deposits and unexplained income discrepancies.
A call for balance between innovation and responsibility
The regulator is not against AI, but wants transparency and accountability.
While the Gambling Commission recognised the value of machine learning in detecting complex risk patterns, it emphasised that automation cannot replace human understanding. The warning is that, AI can support compliance, but not define it.
As the UK’s gambling industry continues to modernise, the regulator expects operators to keep up, not just technologically, but ethically and procedurally.
Source: Focus Gaming News
