Cloud-native legal investigation and litigation platform Everlaw Inc. today announced that it has added new generative artificial intelligence features to its portfolio of tools in a beta program to assist legal teams when writing and reviewing cases.

Litigation is a competitive and complex task for legal teams and can involve reams of documents and data, requiring hours of work and a tapestry of facts spread across hundreds or thousands of sources. Drafting motions and briefs is a time-consuming task, according to research from Bloomberg Law, that requires on average more than 20 hours of work per month for individual litigators.

Everlaw’s platform provides legal teams with a comprehensive view of all the data of a case and their litigation data in one place, including predictive coding and machine learning and case preparation tools.

The generative AI the platform uses is the same technology behind OpenAI LP’s ChatGPT chatbot, which can provide litigators the ability to plumb that information for insights and review by summarizing it for key facts or topics or interests.

The AI can also assist legal teams by building evidence-based legal texts in the company’s Everlaw Storybuilder trial preparation platform that assists legal teams in crafting case narratives and workflows by organizing documents and doing reviews. By using information from stored documents, the AI can create a rough draft from key evidence with full citations that are prepared for editing and fact-checking. A task such as compiling statements of facts that might have taken a legal professional hours or days could be done in a matter of seconds.

“The real value of generative AI is when it evolves beyond a ‘tech demo’ and into the embedded workflows where legal teams live – drafting and reviewing materials, and connecting insights across a corpus to chart a straighter path to the truth,” said AJ Shankar, founder and chief executive of Everlaw.

In order to combat potential issues that might crop up when using a technology such as generative AI, the team focused on designing guardrails to help minimize “hallucinations,” or when the AI confidently states completely made-up facts. To do this, Everlaw’s engineers focused on careful prompt engineering to reduce errors and carefully tailored use to only the documents it would be acting on in order to narrow its focus.

This is important given a recent case where a lawyer from the law firm Levidow, Levidow & Oberman used ChatGPT to produce a brief in an injury case involving an airline and the chatbot completely made up more than a half-dozen case citations.

“The race to push generative AI into production is real,” said Ryan O’Leary, research director at market intelligence firm IDC. “It is impressive to see Everlaw doing it quickly and thoughtfully, adding guardrails so legal teams can gain benefits while better navigating the pitfalls of gen AI”

In addition to the safeguards of the AI model, Shankar said that the AI was specifically tailored to provide specific evidence to clients as to where its information and cite it. That way, at any point, a human can verify the information and fact-check it against the actual documents and case law. Although the assistant is designed to reduce the time-consuming effort of compiling evidence and writing the motions and briefs themselves, it’s not supposed to replace the work of the lawyers entirely.

Shankar added that the platform also keeps the security and privacy of clients in mind and their legal data is not used to “train” the AI model. That means documents and sources will not be shared with any third party or be added to the AI, so there is no chance of proprietary or sensitive information leaking.

“Our strategic approach is delivering all the transformative benefits of AI within our trusted platform built upon the enterprise capabilities that the legal industry deserves,” said Shankar.

Image: Pixabay