HPE Blog, UK & Ireland
Chris_Ibbitson

Enabling regulatory compliance in banking with AI

In my last blog, I shared some thoughts on how different forms of AI could not only help improve the customer experience, but also ensure that the bank remains in compliance with various industry regulations.  As promised, the focus of this blog will be on how AI can both help ensure/monitor regulatory compliance, as well as some of the regulatory considerations when it comes to using AI in financial services.

No matter where in the world a bank operates, regulatory compliance is key.  It is its license to operate – to do business.  Failure to be compliant with relevant banking regulations can lead to fines, reputational damage, and potentially the license to do business being removed.  Firms are subject to a wide range of regulations – dependent on the products and services they offer, and where in the world they operate. 

As I’ve mentioned in past blogs, one of the more commons solutions AI is targeted against when it comes to regulatory compliance is around financial crime.  One key area of financial crime compliance is focused on Anti Money Laundering (AML).  AML regulations are focused on ensuring that financial services firms prevent money gained by illegal activity being “laundered” to appear clean.  AML traditionally used within firms were rules based systems that often generate a high volume of false positive reports.  As firms have looked to move to AML solutions that leverage AI – either commercial packages, or in-house developed, they’ve seen an reduction in both false positives.  This leads to a higher quality of alert being generated for investigation (allowing staff to be more targeted on better alerts), and also improves the customer experience for those customers who aren’t laundering money!

AI can help inprove the quality of alerts AML systems generate for human investigationAI can help inprove the quality of alerts AML systems generate for human investigation

 Another important areas of financial crime compliance, is that of sanctions screening.  Sanctions are financial restrictions but in place by a government to achieved a specific foreign or nation security objective by either limiting the provision of certain services, or restricting access to markets, funds or economic resources.  Sanctions are becoming a more popular tool of choice for governments to leverage in foreign policy – both targeted against countries, along with particular businesses and individuals.  As a result, over recent years, global sanctions regimes have become more challenging for firms to ensure they remain compliant with, and have the appropriate screening processes in place.  Failure to comply with sanctions requirements can be costly – as this FT article highlights with BNP Paribas agreeing to pay a $8.9bn fine in 2014!  Similar to traditional AML solutions, tools used to monitor against sanction infringements have traditionally been rules based.  Again, as firms look to adopt AI based solutions to carry out sanctions screening, we’ve seen an improvement in the quality of alert being generated and a reduction in false positives.  This is achieved by leveraging a blend of two AI technologies - Natural Language Processing (NLP) and Machine Learning (ML) – to process data from both traditional sources and also non-traditional sources.

Looking inwards to internal compliance, many different regulations require the surveillance of digital communications both between regulated staff, as well as between staff and clients, to prevent misconduct.  Regulations such as MiFID II and the Dodd-Frank act all require a pro-active approach to surveillance of digital communication.  Whilst many firms will be leveraging recording / archiving systems for voice calls, email, chat messages etc, many will struggle with pro-actively analysing the data to identify suspicious activity.  Again, NLP and ML can help by transcribing spoken communications and then analysing the data against industry specific rules.  This paper explains more on how NLP can help unlock insight from data in regulated industries.

As a result of the growth of regulatory compliance requirements, we are seeing an emergence of technology companies that are purely focused on helping firms solve challenges associated with compliance – these firms are collectively known as “RegTech”.  The field of capabilities in this space is large – with solutions ranging from helping with AML, know your customer (KYC), Sanctions, through to e-surveillance.  One particular area that is gaining interest from firms is solutions that leverage AI to analysing how regulations compare to each other, where there are difference or consistencies, to enable compliance teams to quickly understand where changes to controls may need to be made.  It’s important to stress that RegTech products on themselves don’t ensure compliance – they merely act as enables to the wider firm controls / compliance teams.

AI can help spot the differences between regulations and join the dots on similaritiesAI can help spot the differences between regulations and join the dots on similarities

 Changing direction, whilst AI can help enable regulatory compliance, as you’d expect it’s use across financial services is subject to significant interest by the industry regulators.  Of particular interest to regulators is the importance of transparency.  As this article by the UK’s Financial Conduct Authority (FCA) highlights, “an important function of transparency is to demonstrate trustworthiness which, in turn, is a key factor for the adoption and public acceptance of AI systems”.  Similar views are also shared by the European Banking Federation, along with the Financial Industry Regulatory Authority in the US.  As firms leverage AI solutions (particular those utilising ML) to make decisions that impact customers or markets, being able to explain how the decision has been made, along with what data was used, is critical in ensuring that bias hasn’t accidently been incorporated. 

As the regulators evolve guidance and regulation on the use of AI, many are looking to leverage the experience of the regulated entities.  A great example of working with industry (both technology companies and regulated entities), is the UK’s FCA and Bank of England have formed a AI Public-Private forum to share information and experience of the adoption of AI and ML across financial services, along with gather views on potential principles or regulations.

Going forwards, the use of AI in financial services is only going to increase.  The role it plays in helping enable regulation be both enforced and achieved will grow, as will the regulators interest in ensuring that its use is fair – both to consumers as well as markets in general.  It’s worth noting that the regulators aren’t just interested in how AI is use – but also how firms use AI to drive insights from data.  More on this in my next blog.


Chris Ibbitson
Hewlett Packard Enterprise

twitter.com/Ibbitsc
linkedin.com/in/chrisibbitson/
hpe.com/uk

0 Kudos
About the Author

Chris_Ibbitson

Chris is a Chief Technologist for HPE, focused on the Financial Services industry. Before joining HPE, Chris has worked at both a Global Systems Integrator, as well as at a Global Bank in a variety of senior architectural roles.

Events
HPE Webinars
Find out about the latest live broadcasts and on-demand webinars
Read more
Online Expert Days
Visit this forum and get the schedules for online Expert Days where you can talk to HPE product experts, R&D and support team members and get answers...
Read more
View all