It's Autumn 2023, and Artificial Intelligence is a high-priority topic on many business agendas across the world.

Regardless of industry or function, business leaders are trying to understand how to harness the power of AI – to use it to help improve their performance, and reduce inefficiencies.

But it's very difficult to embed AI into business process when it is still evolving so quickly; and it is crucial for leaders to ensure that there are controls in place, clarity around its use, and that it is not relied upon too heavily in functions where human interventions are still critical.

One of the most important functions, especially in Financial Services, is Compliance and Anti-Financial Crime. These teams are pivotal in ensuring that customers are kept safe, and that criminals are kept at bay. So, can AI help them do this – and is it failsafe?

AI and AML activities

We know that fraud is a billion-dollar problem across the world, and that anti-financial crime professionals often have huge amounts of data, transactions and trends to analyse, in order to detect threats and risks. This is probably one of the key areas in which AI can really impact efficiency. For example, AI systems can:

• Perform advanced transaction monitoring: AI-powered systems can analyse vast amounts of financial transaction data in real-time, identifying suspicious patterns and anomalies
• Carry out accurate customer risk assessments: AI-driven tools can assess customer risk profiles more accurately by analysing a broader range of data sources, including social media activity and online behaviour.
• Forecast: AI can forecast potential money laundering or fraud activities by analysing historical data and identifying trends

By using AI in this way, anti-financial crime professionals can reduce the amount of time spent on going through the data and finding trends; instead, they can spend more of their time on analysis, producing reports and making informed decisions about risks to the business.

AI and Compliance

Regulations are changing all the time, and one of the key responsibilities of a compliance officer is to make themselves aware of changes, adapt their firm's procedures accordingly, and provide evidence to regulators that they are complying with the rules. AI can assist them in doing this by:

• Sifting through vast amounts of 'unstructured' data such as news articles and regulatory documents to draw out key areas of change or update to laws
• Composing summaries and simplified reports which Compliance officers can use to inform their recommendations to the rest of the business

Using AI in this way will reduce a Compliance officer's time spent on reading vast documents and draw out key points quickly, enabling them to use this information to analyse the changes they need to make internally and spend more time on putting new and effective controls or policies in place.

But does AI ever get it wrong?

Accuracy in AML and compliance tasks is absolutely paramount, so it's very important for Compliance and Anti-Financial Crime leaders to understand that AI is not perfect – and does get things wrong sometimes.

AI systems rely on accuracy of data - they use this data to learn and train themselves. If the input data is incomplete, outdated, or inaccurate, it can lead to erroneous results. Coupled with this is the fact that AI systems can very quickly become out of date, especially when it comes to regulatory information where, as Compliance Professionals know, changes in regulation can happen on a regular basis.

AI systems, although they will surely improve quickly, can be inaccurate in their findings; and when it comes to AML, this can result in false positives or negatives, which could potentially add to the workload of an Anti-Financial Crime professional rather than take it away. AI systems can also 'inherit' bias, depending on the type of information they are learning from – so human oversight is crucial to ensure customers are treated fairly.

And what can't AI do?

AI is what it says on the tin – Artificial Intelligence. The speed and breadth of intelligence within AI systems will likely far outweigh that of a human; but humans possess some critical skills that AI systems don't have, at least not yet. When it comes to Compliance and Anti-Financial Crime, human professionals use the following vital skills to ensure success:

• Critical thinking and judgement; the ability to analyse information in the context of lots of other considerations, and form judgements without any bias
• Decision making; the ability to make well-informed decisions within the context of their own role and business
• Relationships; the ability to network, speak with peers, attend events and learn from these interactions
• Ethicality and Trust; humans are able to build trusting rapport with customers, which machines are much less able to do, as well as ensuring ethicality and fairness are always kept in mind

So, as a Compliance and Anti-Financial Crime professional, you may well find that you are starting to use AI systems more and more in your work – but do so with assurance that you are using it in the right way and for the right things. Compliance and Anti-Financial Crime professionals should embrace AI as it continues to develop, and acknowledge that it will become ingrained in your professional lives in a positive way; leaving you to spend more time doing what you do best.

*********************************

This blog was written in collaboration with Kash Baig, a senior compliance professional whom Twenty84 has had the pleasure of working with for many years. Kash describes himself as a sci-fi nerd who loves Star Wars/Trek and all things superheros; an avid reader who dreams of peace and harmony in a dark world – who with the time left does some compliance work! Kash began his career working in criminal law, before moving into compliance around 2010. He has since evolved in to a self-confessed compliance nerd who loves...life!