To ensure the money that flows through its accounts is clean and legitimate, the global casino industry needs to ensure it has a full picture of its countless millions of customers – AI can play a huge role in achieving this, but it will not be a simple task.
Crucial to achieving this will be combining the technology with the long-standing functions. Betting operators will also have to ensure that AI goes beyond just analysing a transaction, it will need to analyse the behaviour of the people making the transaction.
“A lot of processes around reporting could be automated, but the responsibility cannot be passed onto the system for these decisions,” says Piotr Lisak, MLRO and AML Governance Officer at Kindred, speaking at the Payment Expert Summit in Lisbon, on a panel covering ‘Tech & regulation: Fostering a sustainable AML Strategy’.
Addressing a crowd at the Feira Internacional de Lisboa, Lisak made his case that many regulators, and many stakeholders in regulated industries – both gaming and finance are good examples of these – are reluctant to ‘give free rein’ to AI to make anti-money laundering (AML) decisions.
This is due to rising concerns over responsible AI practices. Betting firms cannot just put sole responsibility for AML on a machine build system. As Lisak put it: “AI is a decision support tool that has to be completely transparent.”
Agreeing with Lisak, Luis Carlos Perez, Chief AML Officer at Lottofy, asserted that technology needs to be smart, intelligent and predictive, to ensure robust AML. This is an area the Malta-based firm has gained considerable experience of.
Many Malta-headquartered companies – with the Mediterranean island hosting a massive local sector accounting for roughly one-tenth of its economy – saw first hand the risks of non-compliance with AML as the country was placed under the Financial Action Task Force (FATF) greylist of markets with AML deficiencies.
Several years later, Malta has been able to remove itself from the greylist, and firms like Lottofy, its AML Officer asserted, work to the ‘highest levels of international standards’, including those set by FATF.
Speaking from this position of experience, Perez believes that AI must be used to conduct behavioural analysis, looking at factors like how long a player spends on a gaming platform.
If operators only focus on transaction analysis, he shares, they will not be able to avoid a lot of criminal activity. Regardless, these capabilities need to be used to support human decisions, not supplant them.
“AI should not replace humans,” Perez shares “It is a game changer tool, and is very supportive to the AML department in this case.”
Responsibility for AML does not just fall on the shoulders of gaming operators, financial institutions or payments firms, however. Regulators, as noted above in the case of Malta, also have a key part to play, as do policymakers.
Zeroing in on the EU, the trade bloc has been moving to coordinate its AML standards lately, such as creating a continental regulatory authority. Kindred’s Lisak, however, believes that harmonisation is still some way away, at least for the gaming industry.
Operators are subject to different regulations across different countries, unlike the EU’s banking networks which are subject to some more top-down regulatory initiatives. Betting therefore has to spend a lot of time looking at differences in regulation, differences in authorities, and questions around the use of AI.
“In the EU we are supposed to have harmonisation on the topic, but this is far from the truth. The devil is in the details,” Lisak remarks.
Moving across to the other side of the Atlantic, Ashley McCulloch, Director Account Management North America at Inspired Gaming Group, offers a US perspective. In her view, there is a lot the US industry can learn by monitoring developments in Europe.
This is largely due to the two sharing one similarity – the US and EU are huge political entities home to various jurisdictions, states in the former and member countries in the latter.
There is ‘so much individual regulation’, McCulloch says, adding that this ‘makes it challenging not just as an operator but also as a provider to be competitive in each market.’
She also shares her views on AI, sharing similar sentiments to Perez and Lisak when saying that the tech must not supplant humans. This is particularly important from a public relations standpoint – the betting industry should not be seen to be cutting people’s jobs in favour of a technological replacement.
Lastly, AI is something that has also caught the attention of regulators on both sides of the Atlantic. The creation of the US and UK’s AI Safety Institutes, the EU’s AI Act and multi-million dollar government funding for the tech in Canada are indicative of this.
Stakeholder industries, whether these be in technology, finance and payments or gambling, should engage with emerging regulations. In McCulloh’s view the ideal regulation should focus on AI’s ideal use case – behaviour.
“When we look at regulations we need to look at what behaviour we create, and whether it is the right behaviour we want,” McCulloh concludes.
