How Thomson Reuters is transforming regulated sectors 

Generative AI (GenAI) has been steadily making inroads into these sectors. This evolving technology has the potential to significantly boost efficiency by automating necessary but mundane tasks such as research and document creation, thus freeing these practitioners to focus their valuable time on delivering high-value insights and expertise to their clients. Despite these benefits, many professionals remain concerned about whether they can harness GenAI’s powerful capabilities usefully and safely. And that’s why Thomson Reuters is focusing its innovation efforts on professional-grade GenAI.

Over 30 years of innovation 

Professional-grade GenAI is grounded in facts and data that those using the tool can confidently reference and cite. It has been tested and authenticated to ensure quality and performance. And it ensures that the data it accesses and analyzes is kept rigorously secure and private. In short, professional-grade AI is designed to provide the reliability that practitioners require to maintain their professional standing and their clients’ confidence. 

Thomson Reuters has been building upon its long history of innovation in the information space. This is a legacy that dates to the 1870s and the founding of predecessor company West Publishing, a producer of law books and other legal content. That human expertise is now being put to work in the development of CoCounsel. CoCounsel is an AI assistant designed to benefit users: to automate the processes of performing sophisticated research and analyzing massive volumes of information for legal professionals (and others). 

Thomson Reuters began laying the groundwork for its work in AI in the early 1990s. With the rise of GenAI and high-profile tools such as ChatGPT, Thomson Reuters undertook research to understand how these innovations might impact legal and other professionals’ practices. Now, the Thomson Reuters Institute publishes an annual Future of Professionals Report that tracks the perceptions, viewpoints, and level of AI adoption within various professions. Wong noted that in this year’s report, a majority of respondents said that GenAI could potentially change or do about 200 hours of work per year per person by 2029. 

The importance of the human element 

Thanks to these reports, Thomson Reuters is also aware of the worries that professionals have about GenAI and the reliability of the information it delivers. This is why the human element is crucial to AI development. 

“I think what's unique about Thomson Reuters is our domain experts,” Wong says. The company has 1,200 attorney editors, along with numerous tax and risk experts. Along with analyzing, annotating, and developing content, these knowledge experts are helping develop and validate Thomson Reuters’ GenAI tools. This work includes instructing the large language models (LLMs) that are the technological foundations of these tools. 

Generic LLMs “have only general knowledge,” Wong notes. “They need to be directed and instructed, and so those experts are critical to part of that. . . . We approach all our editorial work without bias and a commitment to have currency and completeness in the jurisdiction or in the area of practice that we provide support to.” 

Innovation that works 

Thanks to the work of our experts, the data that Thomson Reuters taps in its GenAI technology “does not have hallucinations, it doesn't confabulate or manufacture facts,” Wong adds. “We know that AI systems are not perfect. But it's a big difference between being able to understand something is 99% accurate or 90% accurate or 50% accurate.” 

Testing and authentication to ensure the highest level of accuracy “is critically important.” The tools have “to be more than just knowledgeable in the domain,” he adds. “It has to be designed for the domain.” 

As part of its rigorous commitment to technological innovation, Thomson Reuters has nearly 4,500 engineers and 200 data scientists building its GenAI technology. At the same time, “we have human experts evaluating the outputs and the quality of what the generative AI systems create,” Wong says. For example, “if you try to extract a particular clause from a contract, you want that to be done the same way each time you do it.” 

Adopting a culture of innovation 

Morgan Lewis’s Nihill notes that her law firm’s “culture of innovation” includes “the ability to adapt and adjust quickly.” As Morgan Lewis evaluates the use of GenAI tools in its practices and workflows, “what we really admired about Thomson Reuters was their commitment to legal ethics , responsible AI, and ensuring that the technical components of their tools satisfied our obligations as it relates to clients that have a spectrum of preferences as it relates to using AI,” she says. “At a very high nontechnical level, we wanted to ensure that our input, whether that's our prompts or our documentation, could never be anybody else's output.” 

Additionally, her firm’s attorneys “want to ensure that tools don't have a lot of friction in terms of being able to turn on and be able to quickly get up to speed on,” Nihill says. “What we really enjoyed about the Thomson Reuters suite of products was a sleek interface that relied on very good UX and UI design--something that our attorneys would see as rather intuitive in terms of integrating within their legal workflows.”

It all begins with innovation. By staying open-minded to incorporating new technologies and ideas, while ensuring that the tools you incorporate actually serve you, you can stay ahead in an environment of rapidly advancing solutions and expectations. 

Learn more about how Thomson Reuters’ commitment to innovation can help legal professionals stay at the forefront of GenAI technology and provide modern services to clients now and in the future.