Our Take
Multi-step orchestration matters more than the feature list here: breaking complex legal tasks into visible plans before execution addresses the black-box problem that keeps enterprise lawyers cautious.
Why it matters
Large law firms need AI systems that show their work before executing, especially as California's proposed rule would require lawyers to verify every AI output. The timing positions LexisNexis as compliance infrastructure rather than just productivity tooling.
Do this week
Legal ops teams: audit your current AI citation-checking processes before Q3 so you can evaluate whether embedded verification like Shepard's Verify reduces manual review overhead.
LexisNexis shifts from workflows to agentic planning
LexisNexis expanded its Lexis+ with Protégé platform three months after replacing Lexis+ AI, adding six new components focused on multi-step task orchestration. The centerpiece, Protégé Work, moves beyond the platform's existing 300 pre-built workflows to an agentic framework that creates structured plans before executing legal tasks.
Users can either select specific skills or describe legal goals in natural language. The system routes requests to appropriate workflows, presents a visible plan, then produces review-ready work product. Tasks include contract comparison, complaint analysis, research synthesis, and due diligence (company-reported capabilities).
The expansion adds purpose-built drafting agents for contracts, motions, briefs and deal documents through Protégé Agentic Drafting. Outputs now support standard formats: Word for drafts, Excel for structured findings, PowerPoint for summaries, and PDF for final work product.
Three entirely new features launched: Protégé Workrooms for secure collaboration between firms and clients, Shepard's Verify Trust Markers for citation verification, and Protégé BYOK for customer-controlled encryption keys. The BYOK capability has been deployed in AmLaw 100 firms (per company testing).
Protégé Vault expanded to handle 100,000 documents per workspace, accepting audio, video, images and PDFs with source linking to exact passages and timestamps.
Citation verification becomes compliance infrastructure
The orchestration layer addresses enterprise legal's core AI concern: understanding what the system will do before it acts. Instead of single black-box responses, lawyers see multi-step plans they can approve or modify.
Shepard's Verify arrives as California's State Bar proposed requiring lawyers to verify every AI output. The system flags citations that cannot be verified as existing in LexisNexis sources, though it does not confirm whether cited authority actually supports the legal proposition. Still, embedded verification at the drafting stage shifts citation checking from post-hoc manual review to real-time workflow integration.
The customer-held encryption keys target enterprise buyers who need data sovereignty. Integration with AWS KMS, Azure Key Vault, Google Cloud KMS and HashiCorp Vault gives legal departments control over access revocation.
Platform consolidation strategy faces velocity competition
LexisNexis is betting that authoritative content plus enterprise controls will outweigh the product velocity of newer entrants like Harvey, Spellbook, and DraftWise. The company integrates Anthropic-powered skills alongside its own development, maintaining relationships with Google and OpenAI as well.
The rapid release cycle shows competitive pressure: workflow preview in January, rebrand to Protégé in February, now this expansion. Each builds toward positioning LexisNexis content as the differentiator in an increasingly crowded legal AI market.
For procurement teams, the platform expansion creates integration opportunities but also vendor lock-in questions. The workrooms feature potentially competes with existing collaboration tools like Litera Transact and HighQ, though focused on AI-assisted work rather than transaction management.