AI’s use in transaction monitoring has not always been welcomed, but the market is changing, and it is quickly becoming core to these processes.
This came from a new webinar hosted by anti-fraud technology developer Resistant AI. The panel was chaired by FinTrail director Ciara Aitchison and was joined by Resistant AI CEO Martin Rehak, Kroo Bank non-executive director and chair of risk Serena Joseph and Finom head of compliance Kseniia Kutyreva.
The focus of the webinar was how AI is being leveraged to transform transaction monitoring and how it can reduce false positives, identify patterns and detect criminal behaviour.
Kicking off the discussion, Aitchison asked Rehak what the current landscape is for AI’s use within transaction monitoring.
Rehak explained that the whole industry has rapidly changed its position on AI, moving away from scepticism towards wide-spread adoption. When Rehak’s previous company Cognitive Security (successfully sold to Cisco) first started reaching out to potential customers in 2006 for its AI-powered cybersecurity services, they were met with a frosty reception. He said, “When we started applying AI and machine learning to cybersecurity, the first three customers we ever visited told us no one will ever use AI for cybersecurity, because it’s hard to manage. It’s very expensive to set up and it’s hard to explain.” Fast forward to 2023 and AI has become absorbed as the main tool in a cybersecurity process.
A similar situation is unfolding within the transaction monitoring space, Rehak added. Transaction volumes are ever-growing, and it is becoming impossible to manage the volume of data without the use of AI. Not only that, but criminals are becoming more sophisticated and more talented with how they hide their illicit activity. “You don’t need money mules anymore. You can automate them away in the same way as how you automate people in call centres. You can automate members of the criminal gang.”
The only way to combat the magnitude of transactions and the new methods of criminals is through the support of AI.
Following on from this, Aitchison questioned Kutyreva about how Finom adopted AI into its platform and what the general consensus was within the company. Serving as a fully digital B2B challenger bank that was founded in 2019, Finom is supporting freelancers and SMEs across Europe. With such a wide net of customers, this poses a major challenge for transaction monitoring. As a result, Kutyreva stated the digital bank was always aware AI and machine learning would be crucial to their compliance.
“A rule-based system would not allow us to hit our targets in terms of efficiency or effectiveness and it will not be scalable across the EU. Basically, from the first month of operations, we started exploring AI based solutions.” While there was determination to implement AI, there were still reservations about external factors surrounding its adoption.
These ranged from finding the right provider to making it explainable to the regulators. Getting approval from regulators was one of their most daunting challenges, as many had expressed their lack of faith in the technology within transaction monitoring, Kutyreva explained.
In just three years, this opposition is already dissipating. “The regulator in the Netherlands, declared that they will themselves use AI for the purposes of supervision. So, I think that’s a very good indication of where the industry is going.”
For more insights from the panellists, watch the full webinar here.
Copyright © 2023 RegTech Analyst
Copyright © 2018 RegTech Analyst
