IBM RegTech Innovations

Demystifying AI for risk and compliance

Share this post:

Artificial intelligence (AI) is no longer a new idea. From autonomous vehicles to spam filters, AI is affecting our lives and the way many companies do business. A recent explosion of AI applications is taking place in financial institutions, particularly in the area of risk and compliance.  Some may see this as a no-brainer, but others might not think about AI applications outside the more common uses by banking and financial institutions. Some of the biggest applications of AI today in banking and finance are in front-end customer service like chatbots or virtual agents, or in the securities sector by identifying credit card fraud and helping to detect money laundering/creating digital trust, and lastly in the simple task of automating redundant processes.

So how is AI helping risk and compliance processes? Risk and compliance departments suffer from massive data loads and exhausting regulatory requirements. Think about the large numbers of documents and repetitive processes, mainly in automating legal, that soaks up the time of many compliance and risk departments. But what about AI makes it suitable for the task? AI in the form of algorithms has already been around for nearly 20 years and AI at its root is simply a set of statistical processes. But what really separates AI systems from other types of data analytics is AI’s application of an iterative, learning process, which allows it to adapt inputs and perform tasks more efficiently and often automatically. That still sounds complicated, doesn’t it? Yes, a little, I agree. Look at these three things risk and compliance leaders should consider before taking their first steps toward an AI solution or digging deeper in AI processes.

  1. Understand and define what is available. Iterative, rules-based or explicative? To apply a statistical algorithm effectively, financial institutions must know which one to use, and understand its associated processes.
  2. Pinpoint the problem to solve, and its associated use case. AI has seemingly limitless possibilities, so financial institutions must focus on what is important.
  3. Use the right technology. Data is at the heart of AI, and the spread of new hardware is driving AI’s widespread adoption. But different AI tools require different technology combinations.

More information about AI processes can be found in the report, Demystifying Artificial Intelligence in Risk and Compliance, by Chartis Research with research partner IBM. This report offers a step-by-step guide to understanding AI applications for risk and compliance departments. The report showcases how AI tools and techniques are helping financial institutions manage risk and compliance tasks. Applied to certain process, these AI techniques and tools are changing the way financial institutions conduct business. These techniques include: robotic process automation, speeding up routine tasks and minimizing human error and also text analytics and insights, processing unstructured data, and/or identifying relevant content, negative news, case notes, and more. Another AI technique changing the risk and compliance landscape is entity resolution and network analytics, i.e., determining connections between individuals in order to evaluate risky parties and networks. And perhaps the biggest AI technique involves natural language processing (NLP). This AI application is particularly useful in meeting compliance requirements (NLP, in fact, is one of the most commonly used AI processes).

The demystifying AI report also highlights a number of case studies where AI helped risk and compliance departments. These case studies identify the problem being faced, how AI planned to help, what progress in the AI implementation had been completed, and what AI techniques were being used. The case studies include:

  • Legal and regulatory rules management: legal document extraction at a Tier 1 European bank
  • Trading, research and business-aligned risk and control: analysis of corporate bonds at a Tier 1 US bank
  • Investment compliance: document analysis and rules extraction at a Tier 1 US bank
  • Regulatory compliance: document management at a Tier 1 Asian bank
  • Regulatory and internal compliance: AML and KYC at a Tier 1 European bank
  • Limits management and trading compliance: rules extraction at a Tier 1 US bank
  • Trader surveillance and control: Tier 1 US capital markets firm

Remembering that artificial intelligence is a very broad term with multiple definitions and applications is very important. In order for AI applications to be successful, financial institutions need to connect specific AI techniques to appropriate tasks to deliver results and create actionable insights. Learn more about how IBM artificial intelligence solutions can help banks and financial institutions create value in risk and compliance areas.

Vice President, Financial Crimes & Conduct Risk

More IBM RegTech Innovations stories

Succeeding in the future of payments

Embracing the future in the age of instant gratification The payments industry is at an inflection point. Customers increasingly demand new payment options like real-time payments and P2P payments, underpinned by security and trust. The costs of running payments operations, complying with a myriad of regulations and enforcing payments standards are increasing. Exposure to operational […]

Continue reading

AML is a global issue – are your risk controls global too?

Money laundering is an increasing global issue, often involving multiple countries, intermediaries and organizations. One needs to look no further than recent headlines concerning an Iranian-backed Lebanese terrorist organization or a Malaysian government development company to witness this troubling trend. AML and sanctions risk in every corner of the world As the United States increases […]

Continue reading

Stop money laundering with a proactive AML approach

The recently revealed $8.8 billion Troika Laundromat offshore money laundering scandal was a vast and complex deception. Not just a limited scheme, between 2006 and 2013 it facilitated fake trade deals, reinsurance fraud, tax evasion, hidden investments, a fuel pricing fraud scheme at Sheremetyevo Airport—and much more. The Organized Crime and Corruption Reporting Project and […]

Continue reading