LexisNexis Legal & Professional today (22 July) announced a set of new capabilities in its gen AI solution Lexis+ AI, which enables conversational search and provides summarization, intelligent legal drafting, and document upload and analysis capabilities.
Lexis says that all recent and upcoming enhancements are based on customer-led development.
Usability enhancements include:
- “Stop Response” Option – Gives users more control over the AI assistant’s actions, allowing them to halt responses if they need to make changes to their prompts, which will save them valuable time.
- Default Jurisdiction – Users can now set a jurisdiction as their default jurisdiction, which allows it to appear pre-selected each time they access the jurisdiction panel for seamless workflow.
- History – A new conversation history feature enables users to easily find, review, and re-run previous conversations with the AI assistant.
- Lexis+ AI Uploads Grid View – Streamlines the research process by displaying answers from multiple user-uploaded documents on a single screen for clear side-by-side comparisons and efficient insights.
Legal research and drafting enhancements include:
- Lexis+ AI Headnotes – Uses advanced AI technology to provide comprehensive headnote coverage for every case across the entire collection, allowing users to quickly grasp key points of law and easily locate similar cases, ensuring no critical legal insights are missed.
- Shepardize Uploaded Documents – Users can run the Shepard’s citation service directly within their uploaded documents using the AI assistant. This provides access to the full Brief Analysis report, identifying weaknesses in opposing arguments and strengthening their case.
- Shepard’s Treatment Summaries – Enables users to quickly determine if a case is relevant with at-a-glance summaries and seamlessly access the full Shepard’s report.
- Legal News Integration – Direct access to premium legal news content is now integrated into the AI assistant, complementing existing legal research capabilities for comprehensive legal analysis in one solution.
Retrieval Augmented Generation (RAG) enhancements now include:
- Integration of advanced Shepard’s Knowledge Graph to the proprietary LexisNexis RAG platform (GraphRAG) – Users can harness the power of Shepard’s case law relationship information to deliver more authoritative, complete and final AI-generated responses.
- Shepard’s ® Knowledge Graph is a sophisticated data organization framework that allows users to:
- Precisely identify the most relevant data from extensive Lexis legal sources.
- Receive complete answers based on connections between related opinions, legal taxonomies, and additional metadata that enhance the quality and context of responses.
- Accurately verify citations, case connections and legal concepts as validated by the Shepard’s® Citations Service.
- Reduce research time and improve confidence and trust through measurably better answer quality.
Large Language Models added as part of Lexis’ multi-model approach include:
- Claude 3 models by Anthropic, hosted on Amazon Bedrock from Amazon Web Services (AWS)
- GPT-4o, hosted on Microsoft Azure OpenAI
- Fine-tuned Mistral 7B, an open-source model, hosted on LexisNexis-managed servers within AWS
“We are committed to a diverse and wide set of large language models in the legal space—and the speed at which we investigate new models, experiment with them and deploy them is unmatched,” said Jeff Pfeifer, chief product officer, Canada, Ireland, UK and USA, LexisNexis Legal & Professional. “We are focused on delivering the highest-quality answers to our customers with unparalleled speed and driving trusted results in areas that customers have indicated are their top priorities.”
Lexis+ AI was recently at the centre of a debate around the rate of hallucination after Stanford University published an in-depth research paper that claims that legal research tools such as Westlaw and Lexis+AI hallucinate between 17% and 33% of the time. LexisNexis pushed back hard on the claim and you can see some of their responses here.
LexisNexis announces new capabilities for Lexis+ AI including RAG enhancements
