Thomson Reuters Innovates Legal Research with AI, Accelerating Professionals' Work

Which company's case study is this?

Thomson Reuters is a global leading company providing information services for professionals in law, tax, accounting, and compliance. Building on its vast accumulated expertise and data over many years, it combines cutting-edge technologies like generative AI to develop and deliver solutions that fundamentally transform the way professionals work.

The challenge they wanted to solve

Legal professionals, such as lawyers and corporate legal staff, daily need to quickly find accurate information relevant to specific cases from a vast amount of legal documents, precedents, and regulatory information. This research work was extremely time-consuming and a major barrier to increasing productivity. Traditional keyword searches often yielded many irrelevant results, requiring significant effort to arrive at the truly necessary information.

How AI was used

Thomson Reuters developed "Ask Practical Law AI," a generative AI assistant based on its reliable legal practice content, "Practical Law." This tool takes complex legal questions entered by users in natural language, and the AI searches, analyzes, and summarizes vast professional information within Practical Law, generating concrete answers with cited sources in seconds. This mechanism utilizes Retrieval-Augmented Generation (RAG) technology, which generates answers by referencing only trusted internal data sources, significantly reducing the risk of hallucination (fabrication of information), a common challenge with general generative AI, thereby ensuring reliability that professionals can trust.

Introduction effects and key points to watch

• ▸It dramatically improves the productivity of lawyers and legal staff by reducing complex research tasks, which previously took hours to days, to seconds to minutes.
• ▸By streamlining research tasks, professionals can dedicate more time to high-value strategic work and client interactions.
• ▸Answers clearly state links to the authoritative Practical Law content, allowing users to easily verify the accuracy of the information.
• ▸Notably, by using the company's own vast, specialized, and reliable knowledge as the AI's foundation, it meets the most crucial requirements in the legal field: accuracy and reliability of information.

What Japanese companies can learn

This case study serves as an excellent model for companies with specialized knowledge or unique internal data seeking to leverage generative AI. By building RAG based on their strong suit of reliable data, companies can develop high-value services for customers or tools to streamline internal specialized operations. This approach demonstrates how crucial it is to "ground" generic generative AI with proprietary data, rather than using it as-is, especially in industries where information accuracy is fundamental to business, such as finance, healthcare, and legal services.

Today's AI News

Reference Links