|
24 August 2026 (Nashville, TN) - Legal artificial intelligence platforms are evolving beyond chat-based assistants to operating systems that can operate autonomously with limited user supervision.
LexisNexis Legal & Professional, Thomson Reuters and Harvey have announced new offerings, featuring platforms capable of retaining context, orchestrating multiple AI agents, and connecting to firms' knowledge systems. These announcements suggest that the future is less about AI chatbots for lawyers and more about all-encompassing platforms.
Today, LexisNexis Legal & Professional released new agentic capabilities in Lexis+ with Protégé, in which the platform selects and deploys the models, agents and skills for the user during a project. Agentic AI refers to systems that can act autonomously. Said Sean Fitzpatrick, CEO of LexisNexis Legal:
"Legal professionals should not have to spend their time managing AI, switching among tools or repeatedly explaining how they want work performed. Legal professionals can work intuitively in one experience and receive well-structured, review-ready outputs grounded in sources they can trust".
This new centralized capability stands in stark contrast to earlier iterations of legal AI, in which users would have to prompt an AI chatbot to complete each task within a project and sometimes switch between different AI tools.
It's a trend that has been catching on among major legal AI vendors.
Last week, Thomson Reuters released the next generation of CoCounsel Legal, which included agentic capabilities that handle multiple tasks within a legal project, including research and analysis.
Joel Hron, the chief technology officer at Thomson Reuters, said the evolution of legal AI is part of a broader trend where traditional AI applications that depend on prompting are shifting to agentic AI systems.
Instead of a simple input and output chatbot that connects to an AI model, Hron said that agentic applications are more "goal-oriented," where AI models operate under specific directions to use capabilities and information to solve a problem:
"The prompt is not quite as important anymore because the agent itself is progressively discovering the task along the way".
Even though the agents are working autonomously, Hron added that users must still provide insight and guidance to ensure that AI systems produce accurate results.
The legal AI company Harvey released a new generation of its platform, called Harvey II. The changes mirror a broader industry move from isolated prompting toward AI systems embedded in actual workflows.
Harvey II includes a feature called memory, which can learn and retain how individual lawyers work. There are also spaces built around a matter or project. David B. Cunningham, chief innovation officer at Reed Smith, said:
"The use of memory, personalization, and understanding why decisions were made on past matters are fundamental to our deal cockpit. Harvey II's investment in making these more native to the everyday experience of drafting and negotiation are foundational to our firm intelligence and transactional work".
Hilary Goodier, a partner at Ashurst Perkins Coie, said that while the new Harvey functionality enhancements may seem small, they are important steps along the journey of personalization:
"For that reason, their impact on how lawyers work and the value they derive from the platform shouldn't be underestimated".
| 
