In my past experience, the Legal Departments of large financial services organisations are normally quite risk averse and are not well known for embracing new technology or championing innovation. I was therefore surprised, overjoyed in fact, when I spoke to Kevin Johnson, Head of Innovation Competence Centre at Euroclear, to discover that their first production use case of Generative AI (GenAI) was for, and driven by, the Legal Department!

Kevin went on to share that Euroclear also runs a “legal hackathon” as the legal team are really embracing innovation and design thinking to put themselves at the heart of what they are doing as business transformation, which of course increased my excitement to learn more.

Generative AI in Action: Legal Co-pilot

Euroclear's pioneering generative AI application is the "Legal Co-pilot." This tool is designed to enhance the efficiency of the legal team by automating the initial stages of legal research.

Kevin shared: "Our current view of Generative AI is this is an augmentation to human capability, not a replacement of it. So this will be for our legal team who came up with the Business Challenge. Every time they get questions from the business about whether this regulation is applicable in this situation, they're spending hours going through regulatory watch information that we prepare when a new regulation comes out, to understand if it is or if it isn't."

The Legal Co-pilot addresses this by transforming legal queries into prompts, delivering concise, natural language responses drawn from extensive legal documentation.

Kevin cautioned that they have turned the “creativity” of the language model right down so as to eliminate hallucination as much as possible. In addition, the system not only points lawyers in the right direction but also references the documents where the answers are found, significantly reducing the time spent on preliminary research.

Strategic Implementation: Beyond the Legal Department

Euroclear's approach to generative AI extends beyond legal applications and Kevin discussed the potential to adapt the core technology of the Legal Co-pilot for other departments within the company. He shared that they initially a number of use cases they were investigating, and they decided to start with the legal use case not only because it was the strongest, but also because it would give them the best opportunity to demonstrate the capability of GenAI while building something that could be leveraged by other departments.

Like most financial institutions, Euroclear sits on mounds of data and internal documentation, so he expects multiple applicable use cases where GenAI could play a role, particularly to gain operational efficiency.

One of his key considerations however is “should it?”, cautioning against simply applying GenAI to every area across the business.

This adaptation strategy highlights Euroclear's philosophy of generative AI as an augmentation tool rather than a replacement for human expertise, aiming to boost productivity across various functions.

Regulatory and Ethical Considerations

A critical component of implementing generative AI in financial services is navigating the complex landscape of regulations and ethical standards. Kevin described Euroclear's proactive approach: "One of the first things we did was around the governance. Bringing together people from across the company to actually look at this, and have those conversations that wouldn’t happen naturally." This collaborative effort ensures that any AI deployment is compliant with existing laws and ethical guidelines, integrating insights from Euroclear's legal, compliance, and ethics teams as well as business and technology.

When I asked Kevin if he thought regulation would inhibit innovation, he answered most emphatically: “I'm sitting inside an institution that is regulated by over 20 regulators, national, international, supernational, and the phrase we can't do that because of regulation is not something that we take lightly, and if it really is that we can't do it because of regulation, then there's going to be a very good reason!”.

Kevin views regulation as useful and is quick to point out that if you really look at why regulations are there, they are there to protect end consumers.

The Path to Production and the Role of Partnerships

Euroclear currently utilises off-the-shelf Generative AI technologies to expedite the deployment process while ensuring stringent security measures. "We're currently using off-the-shelf components and assembling them together in our own environments to make sure that the actual security is there," Kevin stated “there is no ‘leak back’ out to the internet in any way”.

This approach underscores the importance of balancing innovation with security, particularly when handling sensitive financial data.

Educational Initiatives and Employee Engagement

Understanding and correctly utilising generative AI is fundamental to its successful integration. Euroclear prioritises educating its workforce about the capabilities and limitations of AI technologies. Kevin notes the importance of education to make sure people understand what it can and can’t do. They have even brought in external speakers to talk to employees and really break things down in simplistic terms so employees can understand how it can help them and how it could be used in their day to day activities.

This educational initiative fosters a culture of informed experimentation and thoughtful application of AI within the company.

Future Prospects and Predictions

Looking forward, Kevin reflected on the broader implications of generative AI, not just for Euroclear but for the financial industry and society at large. He speculates that the ability to effectively prompt AI will become a critical skill, akin to the shift observed during the rise of internet search engines. "A person who knows how to prompt properly is going to be the right person," he predicts, highlighting the evolving nature of knowledge work.

Conclusion

Euroclear's strategic foray into generative AI showcases a thoughtful, measured approach to adopting cutting-edge technology within the regulatory confines of the financial industry. By focusing on augmenting human capabilities, ensuring robust governance, and fostering an educated workforce, Euroclear is not only enhancing its operational efficiency but also setting a benchmark for AI use in financial services.

(Download our white paper "Generative AI in Benelux Banking: Opportunities, Challenges and Future Outlook" for a consolidated collection of insights shared by Benelux bankers from ING, KBC, Rabobank, Spuerkeess, De Volksbank and more, across 1:1 interviews, keynotes, panels and executive round tables that we hosted.)