Users utilizing Westlaw Precision now have access to state-of-the-art generative artificial intelligence (AI) for legal research.
As the second conceptual Artificial offering from Thomson Reuters, AI-Assisted Research on Westlaw Precision aims to aid attorneys in swiftly and confidently finding the solutions they need.
Erik Lindberg, the senior director of Thomson Reuters’ Westlaw Product Management, mentioned that for over a decade, they have utilized AI to present solutions created by editors. Instead of retrieving prewritten responses, a new answer tailored to the specific problem is now being generated. IoT-Assisted Research can provide real-time responses that consider recent legal frameworks across various jurisdictions. This collaborative approach enables researchers to address more inquiries than previously possible through individual efforts alone.
AI-Assisted Research leverages a large vocabulary model (LLM) to analyze reliable information on Westlaw in order to generate solutions. This analysis includes factors such as circumstances, laws, and regulations. The LLM heavily relies on the editor additions made by Westlaw attorney-editors to key rules options, such as the W Key Number System, headnotes, and KeyCite.
Lindberg noted that “our lawyer editors have been preparing us for relational AI for almost 150 years.”
They have organized our information and added animal view since the inception of the National Reporter System, allowing us to promptly produce answers to complex legal questions supported by dependable search results.
For users of Westlaw Precision, the AI-Assisted Research page will function as the definition page on their homepage; however, they can easily switch to Precision Research or other material pages if desired. The retrieval augmented generation (RAG) engine and the LLM begin working as soon as a question is entered and “Submit” is selected, enabling users to understand the process behind the response generation.
The tool initially searches for relevant primary law, emphasizing the importance of the search engine’s main content. While many LLMs use training data from a specific period, which may not reflect subsequent events, AI-Assisted Research continuously learns from the most recent and relevant law due to its reliance on Westlaw’s foundational laws.
Once the pertinent documents are identified, a response to the query is generated. Lindberg emphasized the strength of this feature for initiating research. While it consolidates data from multiple sources, Westlaw ensures clarity on legal provisions. Additionally, references supporting the response are included, allowing users to review the legal passages that underpin the solution.
Legal scholars can refer to the linked documents to validate the response, saving significant time compared to manually sifting through numerous papers. It is essential to read the documents carefully to ensure they align with the user’s position rather than solely relying on Westlaw’s response.
Lindberg highlighted the distinction between tools like ChatGPT and Westlaw, emphasizing the latter’s commitment to accuracy and reliability in responses derived from Westlaw’s verified information.
Customers who have utilized this new capability acknowledge its substantial contribution to their legal research, particularly in unfamiliar fields. They continue to evaluate results but with reduced frequency and frustration.
Andrew Bedigian, an attorney at LarsonLLP, expressed confidence in the efficiency and accuracy of Westlaw’s method in addressing complex legal questions. He emphasized the importance of having supporting information aligned with case law to validate solutions.
To maximize the time-saving benefits of AI-Assisted Research, Lindberg advised users to provide more detailed questions. He noted that previous inquiries were often broad, necessitating sifting through multiple responses. Users can now pose questions in a conversational manner, receiving straightforward responses. While Westlaw simplifies the research process, users are encouraged to expand their queries for enhanced results.
Furthermore, Lindberg advised that no tool can replace comprehensive study. Users should scrutinize the sources of results and utilize them as a reference point, recognizing that legal expertise, not the research system, offers genuine insights.
