Legalweek 2025: Embedded And Agentic Generative AI Expands Software Capabilities
This year’s Legalweek conference in New York City was well attended and dominated by talk of vastly improved generative AI capabilities. The themes shifted from last year’s, which focused on generative AI as a new, more intuitive software interface that enabled legal professionals to access information more easily.
This year, the discussions matured. Across briefings, it became clear that legaltech vendors are no longer just layering AI onto existing products — they’re embedding it deeply into their platforms to surface meaningful insights from the data they already manage. From case management and document review to billing and workflow systems, companies are transforming operational data into strategic value.
A key driver of this shift is the rise of agentic AI — systems that don’t just generate text, but take action, follow multistep instructions, and make data-informed decisions. These agents are enabling platforms to move from passive tools to proactive systems that anticipate user needs, offer tailored recommendations, and enhance capabilities far beyond the original software’s design. This emerging trend was evident in nearly every conversation I had at Legalweek.
For example, in my briefing with DISCO, the focus was on expanding the platform from a traditional e-discovery tool into a comprehensive litigation intelligence system. Their team described how AI is now used not only to accelerate document review, but to identify themes, develop case strategy, and even inform risk assessment == all by leveraging the platform’s existing review and metadata. They also highlighted their investment in empathy-driven design, positioning AI not just as a tool for automation but as a means to support the changing roles of legal professionals.
During the LexisNexis briefing, I learned about Protégé, their new generative AI assistant, which is designed to automate up to 20% of legal work — everything from drafting to clause analysis — by pulling directly from embedded DMS and contract data. The goal is to carefully balance automation and control with agentic AI’s ability to personalize workflows, recommend actions, and dynamically generate content tailored to each user’s practice area or jurisdiction. LexisNexis is relying on its vast dataset to position Protégé as a trusted assistant capable of producing real, usable work product.
Similarly, Epiq’s launch of Epiq AI Labs and its AI Discovery Assistant earlier this year evinces a clear move toward foundational AI infrastructure, not just a simple point solution. The tool is designed to automate over 80% of review processes and complete them up to 90% faster than TAR. But what’s even more interesting is how they’re architecting this to support broader service delivery by combining proprietary tools, Microsoft and AWS partnerships, and custom development capabilities, building AI that taps into litigation and review data to deliver actionable insights.
During my conversation with Onit, it was clear they’re betting big on AI agents as a core part of their roadmap. These agents will soon be able to execute complex workflows, improve task accuracy by 20%, and operate across multistep legal and business processes. What struck me was how Onit sees this evolution as a natural extension of their existing data-rich platform. By embedding AI into workflow and billing systems, they’re enabling legal ops teams to surface trends, automate decisions, and ultimately focus on higher-value work.
Clearbrief’s briefing stood out for its focus on pro bono work, but it also reinforced the broader trend: embedded AI tools using existing legal documents to provide real-time, actionable support. By applying generative AI to briefs, filings, and asylum case materials, Clearbrief helps lawyers improve accuracy, reduce administrative load, and increase impact — especially in resource-constrained contexts. Their work shows how strategic value doesn’t always mean revenue: in this case, it means reach, access, and systemic improvement.
Servient’s team talked through their long-term investment in machine learning and how it laid the groundwork for today’s AI enhancements. What’s new is how they’re using those systems to synthesize facts, generate research plans, and draft outlines using litigation data already housed in their system. They’re also exploring agentic models that can operate securely within firm-specific environments, helping firms maintain control over their workflows while gaining intelligence from their own data.
NetDocuments is going deep on embedded AI, particularly through its NDMAX Studio, which lets users create custom legal apps that leverage the data already in their DMS. The platform’s semantic search, auto-classification, and analytics tools are all built to surface governance, productivity, and business intelligence insights that were previously buried in unstructured content. Their team also discussed the vast potential of agentic AI workflows that allow users to delegate multistep tasks to AI — further supporting the theme of data-informed automation that moves legal work forward without manual oversight.
The themes that emerged from Legalweek 2025 this year highlighted how generative AI is no longer just a front-end interface — it’s becoming the engine that drives strategic value by leveraging the data already stored in legaltech platforms. From case strategy and contract analysis to billing and matter trends and document governance, companies are embedding AI to transform their systems from passive tools into active participants in legal work.
Agentic AI is accelerating that shift, enabling software tools to automatically complete tasks and generate outputs that are increasingly indistinguishable from human work. The platforms that succeed will be the ones that make the most of what they already have: rich, structured data — and AI smart enough to use it.
Nicole Black is a Rochester, New York attorney and Director of Business and Community Relations at MyCase, web-based law practice management software. She’s been blogging since 2005, has written a weekly column for the Daily Record since 2007, is the author of Cloud Computing for Lawyers, co-authors Social Media for Lawyers: the Next Frontier, and co-authors Criminal Law in New York. She’s easily distracted by the potential of bright and shiny tech gadgets, along with good food and wine. You can follow her on Twitter at @nikiblack and she can be reached at [email protected].
