The rapid evolution of technology turned regulatory compliance into a daunting frontier. Firms are not only required to keep up with changing technologies, but they also need to stay on top of increasingly complex requirements. Priya V. Misra, who sits at the forefront of this arena, is a pioneer in using LLMs to do the heavy lifting when it comes to compliance.

We recently spoke with Misra about his latest venture, the current regulatory environment, the landscape of LLMs, and advice for leveraging GenAI tools.

Tell us about EKAI and the problem you are trying to solve.

Priya V Misra: EKAI is the first AI compliance ‘co-worker’ for risk and finance professionals. The newer regulations are more complex and reporting on them is more frequent. Compliance platforms of the past are not able to cope with these new kind of reporting requirements. As a private SaaS platform, our proprietary AI software addresses the newer compliance requirements with ease. Our platform is designed to support the compliance plan development and sit at the holistic level for the Chief Compliance Officer and managers to digitally manage their compliance and prepare for regulation engagements and compliance submissions to the regulator. We are currently focusing on newer regulations like Operational Resilience, Consumer Duty, and ESG.

EKAI provides a natural language chat interface to make it easy to use. We believe that as the nature of data has evolved, compliance professionals in financial services also need enhanced systems and tools for this new normal to enable them to support the business and meet the requirements of regulators in a cost effective way. We tested the model with the United Kingdom’s Financial Conduct Authority as part of their sandbox and found regulator alignment for a tool such as our own to aid the industry in meeting their compliance requirements.  

Banks and fintechs have always faced regulatory challenges. Why have concerns around regulation been so heightened in the past few months?

Misra: The concerns around regulation have been heightened in the past few months due to a fundamental shift in different kinds of regulations. The newer regulations are moving from ‘what’ to ‘how’. This basically means that organizations need to show the evidence of compliance, the ‘how’ they are meeting the compliance and not just ‘what’ their compliance approach is. The regulation horizon we see is placing consumer protection at the heart. This could be multi-jurisdictional with the advent of the Consumer Duty regulation in the U.K., to follow similar types of regulations in Europe. For the industry, we see the impacts on costs from being able to handle the multiple and, at times, competing priorities between regulations and maintaining business-as-usual in an environment still competitive for talent. Tools like EKAI offer compliance professionals in financial services better oversight on how they are performing from a delivery perspective against the compliance requirements across multiple programs. 

How have you seen the conversation around GenAI and LLMs evolve in the financial services industry in the last year-and-a-half?

Misra: The conversation around GenAI and LLMs in the financial services industry has evolved in the right direction. The initial trepidation has now been met with wider adoption, with at least of ChatGPT opening the doorway for productivity enhancements from more business intelligence software like EKAI sitting at the intersection of compliance. EKAI has pioneered the use of the Small Language Model (SLM) for AI in corporate usage. SLMs de-risk AI implementation and provide a way to progressively deploy features. SLMs are eco-friendly as they require lower GPU usage and are quicker to train.

Apart from EKAI’s application of GenAI, what is the most powerful application of GenAI for financial services you have seen?

Misra: One of the most widely implemented GenAI applications within financial services that I have seen is in customer support. The power of GenAI lies in its ability to use company-specific information for answering customer queries and intelligently switch to a human counterpart for sensitive queries.

How do LLMs compare to traditional methods of regulatory compliance and risk assessment in terms of efficiency and accuracy?

Misra: The traditional methods of regulatory compliance work for traditional regulations. The data for these regulations was structured and mainly consisted of numbers. LLMs and GenAI are critical in compliance moving forward. They can handle unstructured data of documents, messages, and transcripts. This gives the organizations a strong foundation to build and use compliance platforms.

What advice would you offer firms who are avoiding GenAI tools because of regulatory and compliance concerns?

Misra: Every industry will be impacted by GenAI and/or LLMs eventually. I would advise them to embrace it selectively because it is coming anyway. The GenAI Act coming into force in Europe in the upcoming months will transform the landscape from a sort of ‘wild west’ into one with the types of benchmarks and controls that will ensure its wider and confident adoption across industries in a way that an industrial revolution is supposed to, transforming skillsets and producing efficiency gains future generations should benefit from.

Photo by Markus Winkler