Legacy banks have been historically slow to adopt technology. This is changing rapidly today, as intense competition has compelled these banks to overhaul their dated technology. This overhaul will have a profound effect on regulatory compliance, a dusty but vital corner of investment banking and capital markets.

Challenges of tech adoption in compliance

We are at a crossroads where banks’ compliance back offices are concerned. On the one hand, many have old-school, end-to-end, human-centric compliance processes. The organizational structure and culture of those compliance back offices are shaped by legacy banks’ longstanding practices that make change onerous.

On the other hand, we have modern regulatory technology — or regtech — focused on software solutions for compliance processes but which have not fully integrated automation and algorithms with human decision-making. The challenge is to harmonize technology with a “human-in-the-loop” compliance solution that blends a tech-first approach with human decision-making. Compliance officers immediately benefit from process automation in simply being able to access the information they need most efficiently, setting them up for greater success as these services effectively “learn” from the informed decisions they make.

The solution

This human-in-the-loop compliance solution leverages humans and automation to create machine learning (ML) models that evolve and become more sophisticated and efficient over time, with increasing levels of quality and decreasing error rates. In a regulatory compliance environment, keeping human-in-the-loop processes instills confidence in automation and ML algorithms.

One key example of this approach in action is how data gathering integrates with human decision-making functions for know your customer (KYC) and anti-money laundering (AML) vendors leveraging automation tools. Another is integrating the experience of compliance officers with software workflows to scale operations in a way that is auditable and predictable. The opportunity is to organize compliance touchpoints around an assembly line of information gathering, no matter the individual compliance function, as each workflow is critical to a compliance officer’s performance.

As regtech providers examine and learn about the decision-making of compliance officers, they will deliver software automation with human-in-the-loop processes. Software algorithms won’t make decisions for compliance officers, but they will render more accurate recommendations on the basis of more rapidly and efficiently processed information.

Delivering trust

The world has undergone critical changes over the last two years and will continue to evolve. These seismic shifts will have large consequences even in the more niche domain of regtech.

A harbinger: the Financial Industry Regulatory Authority (FINRA) recently announced its amenability to standardizing remote — that is, virtual — supervision. That is a giant step away from the routine of traditional regulatory supervision — slogging work from city to city, bank to bank, accessing file cabinets in secure data rooms, asking questions in person, etc., but the core information gathering and processing function remains. FINRA’s monitoring will begin, and end, in the cloud.

Times and practices may change, but financial institutions must continue to deliver trust to customers and regulators. Achieving and maintaining that goal demands a blended technological solution — a tech-first, human decision-making approach — that is replicable, auditable and scalable.

The advent of algorithms in the regulatory compliance space doesn’t mean the end of human involvement. In fact, the trust requirement means acknowledging — even highlighting — the importance of keeping a human in the loop. The difference between financial services providers that will flame out and ones that will thrive lies in the adoption of technological processes that still need overt human supervision.