As we discussed in Part I of this blog series, global financial services firms are struggling to keep up with the sheer volume, velocity and variety of regulatory change forced upon them today.
Traditional manual processes are putting a huge strain on the limited, highly skilled in-house resources needed to undertake related compliance tasks.
This adds a huge cost to human-intensive classification and impact assessment, whilst simultaneously increasing the risk that some changes will end up falling between the gaps.
The result is that many organizations end up duplicating work, and fail to generate a unified view of the impact of regulatory change on the business. Blind spots can lead to increased risk and a potential impact on the bottom line. So, what’s the solution?
Time to believe the hype?
Artificial intelligence (AI) is mentioned so often in any discussion of emerging technology innovation that it’s easy to dismiss as a buzzword. But that would be a mistake, as it’s already having a major impact on the financial services industry. AI-powered virtual assistants and chatbots are adding value and reducing call center costs in customer-facing banking environments, for example. Some banks are already using it to drive process efficiencies: BNY Mellon recorded an 88% improvement in processing thanks to robotic process automation (RPA). And innovative start-ups are emerging to tackle serious cybersecurity and fraud challenges with the technology, for example in money laundering and online threat modelling.
The push to embrace new technology platforms is also being led by the regulators themselves. The UK’s Financial Conduct Authority (FCA), for example, has stated it is actively encouraging RegTech solutions “as they could benefit consumers and the wider industry.” Similarly, in June 2019 the US Commodity Futures Trading Commission (CTFA) launched LabCFTC Accelerator to “to facilitate market-enhancing innovation, inform policy, and ensure that we have the technological and regulatory tools and understanding to keep pace with changes to our markets.”
AI for RegTech
So what technologies are we talking about specifically? AI, after all, is only an umbrella term, including:
Semantic web to determine relationships with different types of data in order to automate impact assessment, workflow and the routing of info to certain people.
Natural language processing (NLP) and named entity recognition (NER) which help to bring semantic web technology into the content world, by extracting key stats and topics from huge documents in just seconds and align them to business impact.
Machine learning and deep learning: elements of this are starting to be leveraged across RegTech, although it’s still in its infancy. When mature machines will be able to learn, analyze and assess regulatory change as a highly trained compliance lawyer would.
Speech pattern recognition: could come to the fore in helping to monitor, understand and interpret phone calls and recorded voice messages in the context of trade and market surveillance.
Robotic Process Automation (RPA): can be used in combination with semantic web, big data and machine learning to take human processes and make them faster and more efficient.
Rule extraction and compression: offers fantastic potential in enabling organizations to extract specific controls, rules and logic from a particular regulation so that they can adopt any obligations wholesale rather than having to go through an expensive, time-consuming applicability and impact assessment process.
The key, of course is in using and combining these technologies — or at least the ones that are mature enough — in order to overcome regulatory change management challenges presented by traditional tools and manual processes.
In the final part of this blog series we’ll take a look at how this can be done.