With financial complexities rising and customer expectations skyrocketing, the challenge of trimming excessive compliance has emerged as a vital area for innovation.
“For the past two decades, compliance has developed various systems to identify risks,” stated Green Light AI’s CEO Will Lawrence during a conversation with Pymnts CEO Karen Webster. “But what happens when those systems fail? That’s when people have to step in, which can be a very labor-intensive process.”
According to a cited study, Lawrence emphasized that around 85% of compliance investigators’ work isn’t truly analytical. Instead, it often involves document processing, follow-ups, and other administrative tasks.
“It’s concerning to think my sales team only spends 15% of their time selling,” he noted. “Yet we seem to accept that in compliance.”
Lawrence expressed confidence that, following a $15 million Series A funding round, Greenlight AI has cracked the code on the future of compliance. Instead of merely adding more software to combat financial crime, the focus is on employing smarter AI agents to automate repetitive tasks.
Building in a Highly Regulated Environment
Greenlight’s ambitions aren’t limited to just fintech disruptors. The real challenge lies in serving established financial institutions that are pivotal in shaping market standards.
“You’re a new player dealing with tightly regulated financial institutions. How do you cultivate trust with agents that people might not fully understand?” Webster challenged Lawrence.
As part of their Series A announcement, Lawrence was excited to share that they developed what they call a “trust infrastructure.” This system weaves U.S. federal banking regulatory guidelines into the foundation of each agent.
This infrastructure incorporates model governance, validation, and human oversight in decision-making processes.
The AI-driven solution from Greenlight won’t entirely replace existing compliance frameworks but rather layers AI agents atop them. This aims to alleviate the operational team’s burden of laborious reviews. As Lawrence noted, these agents serve to facilitate required “enhanced due diligence” for high-risk clients.
This shift offers enhanced research capabilities, potentially transforming compliance challenges into competitive advantages.
“High-risk customers could actually be some of your most profitable,” Lawrence remarked. “Yet, they require significant resources, and if you can’t manage that, you might lose out altogether.”
But what exactly makes a customer “high-risk”? It’s not a straightforward answer.
“Common risk indicators include geography, industry, and organization size,” he explained. “But often, it’s about the operational effort required to manage that customer.”
The Evolution of Financial Crime Detection
In one instance, a client of Greenlight AI found that the system was inadvertently flagging a large number of Pakistani users due to name matches on a sanctions list.
The goal, he asserted, is to modify that dynamic while easing the process for legitimate users.
Webster voiced a common frustration: “You’re trying to access a system, and it flags you. But what about the frequency of your visits? These systems should improve detection and mitigation.”
Lawrence emphasized that this is where AI can excel, as it can manage unstructured data and make context-aware decisions to reduce frustrations for genuine users while speeding up compliance reviews.
Reflecting on the timeline of compliance evolution, Lawrence noted that in the 2000s, systems were primarily rule-based, while the 2010s leaned into machine learning. But what about the 2020s?
“We’re entering the age of compliance agents,” he said.
This change is significant, especially for regulated financial institutions where trust is paramount. Errors can lead to not just transaction declines but exposure to regulatory risks.
“AI might seem intimidating until it’s demystified,” Lawrence concluded. “It’s ultimately just a tool, similar to a calculator. Our mission is to help banks navigate its safe utilization.”
