I wrote an article in January last year titled “Only human.” I spoke about how there needs to be a greater understanding and investment into human risk. I was not expecting the growth and evolution of Artificial Intelligence (AI) to be so profound in just 18 months. I was also not expecting the impact AI was going to have on our industry and the world of compliance. As a fun little challenge, I asked an AI tool to write this article… I got a typical list of topics in number order followed by “Conclusion.” We have all seen this on social media posts or worse, in CVs or job applications. So I refused to use it; here’s why…
The many benefits of AI
I won't bang the drum we have all been hearing for the past 12-24 months on how Artificial intelligence (AI) is at the forefront of revolutionising industries and how the compliance sector is no exception. However, as iGaming operators navigate an increasingly complex regulatory environment, the adoption of AI offers new opportunities but also poses significant challenges. Managing this evolving landscape requires an informed and balanced approach, where the strengths of AI are harnessed without overlooking the importance of human judgement.
I recently completed my first course in AI Compliance and have been working with a few organisations to support their implementation of various AI-based tools. One particular company in the financial services industry wanted me to run an audit to test some AML customer reviews that had been completed by their team, or so I thought. In fact, they had been trialling an AI tool for their KYC process and CDD, but I did not know this until after I submitted my report. Ironically, within the first paragraph of my report it read “several of the reviews completed appeared too robotic, ticking boxes but not demonstrating an understanding of the customer profile and the information the company held.”
In April, an operator I work closely with had begun using AI to support their customer service team. The impact was a significant reduction in generic queries reaching the team and a reduction in negative reviews. Does this mean AI was better than humans at the role of customer service or, in my first example, that humans are essential to the AML journey? I think it is a bit of both.
AI’s ability to handle real-time monitoring is transforming compliance from a reactive function to a proactive one. Instead of relying on periodic audits or after-the-fact reporting, AI-driven systems provide continuous oversight
Getting through the data
The use of AI is based on certain promises, such as efficiency, accuracy and scale of operational teams. One of the most significant advantages AI brings to compliance is automation. Tasks that once consumed considerable resources – such as customer due diligence (CDD), Know Your Customer (KYC) procedures and monitoring large volumes of transactions – can now be managed more efficiently by AI systems. I won’t mention any specific providers, because none of them paid me a good enough sponsorship fee to mention their products 😊 (just kidding).
These AI systems process massive amounts of data quickly and accurately, identifying patterns and anomalies far beyond human capacity. Therefore, for those of us that work with a large customer database and a smaller-than-desired workforce, utilising the right AI tool could be a life-saver.
Moreover, AI’s ability to handle real-time monitoring is transforming compliance from a reactive function to a proactive one. Instead of relying on periodic audits or after-the-fact reporting, AI-driven systems provide continuous oversight. This means issues such as potential fraud, money laundering, or irregular betting behaviour can be flagged and addressed in real-time, giving businesses an edge in avoiding regulatory violations.
I work with a lot of operators that have made use of such tools specifically for safer gambling purposes and the feedback has been incredible. As well as the positive impact for their teams, in terms of being able to focus purely on the highest-risk customers, levels of customer safety have increased. We have to consider this a huge win for our industry. I am not advocating for every operator following others and utilising AI tools for safer gambling benefits, because I also work with operators that have invested heavily into their own in-house tools and developing their teams… perhaps the impact of AI is purely healthy competition.
Furthermore, AI is increasingly being used for predictive compliance. By analysing historical data and applying advanced algorithms, AI systems can predict future compliance risks, helping organisations stay ahead of potential regulatory changes or breaches.
It sounds like I am pro-AI and I am giving a great sales pitch… but I actually don’t want to be replaced by AI. So I will talk about the risks now!
The risk of bias & accountability
We should treat AI almost as a child, letting it learn for itself and take decisions, but with someone keeping a close eye on them to make sure it doesn’t stick a fork in a plug socket
I should start with the obvious, such as bias, accountability and over-reliance. While they seem simple, and perhaps risks that could be managed easily, I have seen the contrary. One of the most pressing concerns is bias within AI systems. AI algorithms learn from historical data, which can sometimes include biased or flawed information. This can lead to decisions that unintentionally perpetuate discrimination or unfair practices – particularly concerning when compliance processes touch on areas like player exclusion, responsible/safer gambling measures or financial assessments. So the risk of bias having a negative impact is not one to be overlooked, and perhaps this risk alone is enough to avoid the implementation of an AI safer gambling tool.
In addition, AI lacks the nuance and judgement that human compliance officers bring to the table. Compliance is not always black and white; it often requires context and interpretation that goes beyond what data can provide. An over-reliance on AI systems risks overlooking key human elements, such as ethical considerations or contextual factors that influence decision-making. I love being in an office and hearing members of the various teams such as AML and SG discussing a complex case; bouncing ideas off each other and challenging ideologies. This nearly always results in learning, better processes and a well thought out conclusion. I would never want to lose this within compliance.
Accountability is another challenge. While AI can automate many tasks, it is crucial to remember that responsibility for compliance remains with the organisation. If you sit in a key function position, hold a PML, lead a team or simply work in the day to day of customer reviews, you are accountable for your actions.
In the event of regulatory breaches, businesses cannot hide behind the claim that “the system made a mistake.” Therefore, maintaining a clear chain of accountability – where AI augments rather than replaces human oversight – is essential.
I love being in an office and hearing members of the various teams such as AML and SG discussing a complex case; bouncing ideas off each other and challenging ideologies. This nearly always results in learning, better processes and a well thought out conclusion. I would never want to lose this within compliance
Human supervision
Striking the perfect balance is where I believe our greatest challenge sits. I believe we need both AI and human oversight to achieve effective compliance. We should treat AI almost as a child, letting it learn for itself and take decisions, but with someone keeping a close eye on them to make sure it doesn’t stick a fork in a plug socket.
Given the complex nature of compliance, a hybrid approach that combines the strengths of AI with the wisdom of human judgement is essential. From my perspective, this provides a great recipe for compliance success. To avoid potential pitfalls, organisations should implement governance structures that ensure AI systems are transparent, auditable and free from biases. Ideally, they would appoint someone to lead on all AI implementations and their ongoing management.
Regular auditing of AI models to ensure they are functioning as expected, alongside continuous training and updates to reflect evolving regulatory requirements, should be a mandatory step for anyone considering shifting to a more AI-focused process or systems.
Moreover, operators and B2B providers must establish clear protocols for when AI systems flag potential issues, ensuring human compliance officers review and validate these findings. This allows the compliance function to benefit from AI’s speed and scalability while retaining the depth of human understanding.
The regulatory aspect
I want to briefly touch on the regulation, or lack thereof in this case, and how operators, financial services and B2B providers can succeed in navigating their way through.
The regulatory landscape for AI remains largely undefined, creating challenges for organisations trying to navigate this emerging space. Currently, many jurisdictions do not have specific regulations governing the use of AI in compliance, leaving businesses to operate in a grey area.
To mitigate these risks, companies should stay informed about upcoming regulatory developments concerning AI and ensure their internal policies are robust. Implementing best practices from other industries where AI is more established – such as finance or healthcare – can also provide valuable insights. I work closely with several businesses that have taken AI compliance on a roller coaster of a journey, so there is a lot we can learn from other industries.
The future of compliance is not one of AI versus humans – it is one of collaboration and oversight
In the iGaming industry, however, where compliance is already highly complex, AI must be deployed with care. Over-reliance on AI without clear guidelines can lead to unintended consequences, such as regulatory breaches or reputational damage. As AI continues to evolve, its role in compliance will undoubtedly grow. Its ability to process vast amounts of data, automate tasks and provide real-time oversight positions it as a game-changer for iGaming operators.
Collaboration and oversight
However, businesses must approach AI adoption with caution, ensuring that its benefits are balanced with the critical need for human oversight.
I use the phrase of having a “silver bullet” quite a lot. But AI is not a silver bullet, rather a powerful tool that, when integrated thoughtfully, can enhance compliance efforts and help businesses navigate the increasingly complex regulatory landscape.
By investing in a hybrid approach that combines AI’s efficiency with human judgement, iGaming operators can stay ahead of compliance challenges while minimising risks. The future of compliance is not one of AI versus humans – it is one of collaboration and oversight.
Henderson is the Group Head of Compliance for AMLGS currently, applying his services to GIMO and NetBet Enterprises, who started his career working within UK Law Enforcement Agencies as a Financial Investigator and AML expert advisor.
He has more than 15 years experience advising companies in the UK and internationally within sectors including financial services, hospitality and gambling operations. Henderon is a dedicated advocate for ethics, risk management and inclusion and has set out his strategic agenda for a culture of compliance across the organisations he works within. He is a frequent guest speaker at ethics, diversity and compliance events with the focus on empowering employees within their organisations to achieve their goals and the highest of compliance standards.
