Money laundering has grown rapidly as regulation and bank processes struggle to keep pace with the rise in criminal innovation. And whilst digitalisation – accelerated as it has been by COVID-19 – has enhanced the customer experience, it has also helped create a rapidly widening opportunity for fraud, with criminals able to exploit gaps between digital solutions, banking systems and customers.
To help interpret what’s happening and how the industry should respond, Banking Circle commissioned research of 300 senior decision-makers in European banks as well as conducted interviews with industry experts.
Some key issues identified include:
- How Artificial Intelligence and Machine Learning can enhance the approach to AML
- The role of data quality
- Success will come from adopting an approach that complements, not conflicts with, the business of banking
How Artificial Intelligence and Machine Learning can enhance the approach to AML
When discussing how AI can enhance the approach to AML, Dave Burns, Chief Revenue Officer at Napier, said “Banks are trying to raise their base levels of risk capability, but it’s a question of balancing cost and risk.” Oracle’s John Edison agreed, explaining that “budgets are getting tighter across the board, and driving the balance between investment and reward when it comes to AI will be tricky.”
Despite falling budgets, Banking Circle’s interviewees believed that AI and ML represent the only option for banks to battle money laundering as they transition to a digital future. Their respondents envisaged a future in which robotic process automation is standard, automatically applying ML techniques to data harvested across the entire transaction chain, rather than just select parts of the process as is the case at present.
Chris Caruana, VP of AML Solutions at Feedzai, said AI will “enable us to think ahead to the problems of 2025, using deep learning and neuro-analytics to think like criminals do, especially at the placement stage.” Rene Kartodikromo, Director of Deloitte Netherlands’ Financial Services practice, confirms: “We need to switch from traditional monitoring to smart monitoring – at present, many banks’ AML systems are generating a large amount of false positives.” If anything, false positives are set to get worse, as proposed increases in the contactless transaction limit make higher-value fraud a greater possibility, according to a new report from Featurespace.
The role of data quality
Recent pan-European research of the infrastructure challenges banks face, commissioned by Banking Circle, shows just how big a problem legacy IT infrastructures are across the continent. The survey quizzed bankers about their main IT priorities for 2021.
- 29% – Respondents highlighted data security as their leading concern
- 29% – Creating a better customer experience
- 24% – Improving the quality of data employed
- 23% – Improving AML processes
Professor Brigitte Unger, Chair of Public Sector Economics at Utrecht University, says “many banks are hampered in their use of customer data because of GDPR rules – as are the police.
“The recent public/private partnership between Dutch banks is a model for improving information exchange in the fight against money laundering.” Across Europe, regulators appeared to be taking heed of Professor Unger’s comments. A report commissioned by the European Parliament recommended blacklisting organisations known to have originated money laundering transactions, limiting the role of shell companies and improving data about transaction originators and recipients through national company registers.
Success will come from adopting an approach that complements, not conflicts with, the business of banking
While many organisations like the idea of collaboration and may recognise the benefits, there’s still some hesitancy about getting involved. As Samantha Sheen from Stripe puts it, “many organisations are reluctant to give away their data sources – it’s like McDonald’s’ secret sauce recipe. The problem is not the will to collaborate – external regulators must give clear, definitive guidance on AML, and organisations must be clear that they can share [client] data without violating GDPR rules.”
There are signs that things are changing within the industry, too. Oracle’s John Edison noted that “there are sub-groups where financial institutions are coming together, like Visa, Mastercard and SWIFT, and opening up their API structures to collaborators. We are going to see a lot more of this in the next two years.”
The research concluded with six findings:
- Organisational design matters – As well as implementing new technologies, banks should be breaking down walls inside their organisations and hunting down relevant pools of data across different functions. AML should be elevated within the organisation to be the direct responsibility of the senior risk manager and pull together experts from compliance, operations, technology and risk.
- Introducing Artificial Intelligence and Machine Learning to one’s AML strategy is non-negotiable – With the advent of Faster Payments and, from January 2021, blockchain-based international currency transfers, monitoring and dispute resolution systems that rely on manual processes will no longer achieve the desired results.
- AI-led solutions must be applied consistently, end-to-end, across the value chain. Unless AI solutions and process automation are employed end-to-end, manual-led processes will continue to introduce errors and inconsistencies – as well as slowing transactions down. Financial institutions also need to improve the quality of data they’re using from both internal and external sources, and consider widening the range of sources they use.
- It’s no longer just about banks. Bank clients and partners need to be fully engaged in the battle against money laundering. Regulators are quite rightly placing SMEs and FinTechs under the same level of scrutiny as financial institutions, and banks should be working with clients and partners to ensure AML monitoring and reporting are up to scratch.
- New approaches to regulation are needed –more flexible, more collaborative. Rules-based approaches are going to be increasingly difficult to defend given the rate of change we’re currently experiencing in the market. The answer is a more flexible approach to regulation which includes experimentation (sandboxes) and direct collaboration with banks, fintechs and SMEs.
- Cross-industry collaboration must get better. What Dutch banks are working on in partnership
with the Netherlands’ AFM and consulting firms shows what can be achieved. Banks must now
consider both national and international collaborations, sharing data and approaches to combat increasingly sophisticated and international criminal organisations.