Lexis+ with Protégé AI review (Oct 2026): agentic research and drafting, Workrooms, Shepard's citation checks, Vault analysis, and Claude legal skills

What this Lexis+ with Protégé page covers

This page is for an associate, research attorney, knowledge-management lead, or managing partner at a firm on Lexis or evaluating it. It is educational, not legal advice; memos, drafts, and Vault findings are drafts until a lawyer has read the sources.

LexisNexis announced general availability on 24 Feb 2026 as the platform that replaces Lexis+ AI, with a single prompt box, pre-built workflows (for example a motion to dismiss or discovery documents), a no-code workflow builder, and general models from Anthropic, Google, and OpenAI. A 7 May 2026 release added skills, agentic drafting, Workrooms, Shepard's Verify, a larger Vault, and customer-held encryption keys; a 24 Aug 2026 update carries context from research to drafting and picks a model by default while letting users choose. LexisNexis cites 200 billion documents with four million added daily — its figures.

Ground rules: a licensed attorney owns any filing, citation, demand, deadline, client message, or discovery response that comes out of Lexis+ with Protégé. The associate checks every authority's Shepard's status and pinpoint, reads the case, and only then lets a memo or draft leave the matter.

Major features (with how someone uses them)

Examples at a glance: each feature has a concrete example from a real job plus a short Try this you can run this week.

1. Agentic research and drafting from one prompt (primary)

What it is: You describe a legal goal in natural language or pick a skill; Protégé routes the request, shows a structured plan, and produces review-ready work such as a research memo (7 May 2026). The 24 Aug 2026 update says context carries forward from research to analysis and drafting, LexisNexis legal sources stay the default grounding (web and your own documents only when selected), and answers link to citations with Shepard's signals.

A memo that reads well can still rest on a pinpoint that does not say what the memo claims. Compare against work your firm already knows is right before trusting it on new issues.

Concrete example (associate at a 25-lawyer litigation firm) An associate asks Protégé for a research memo on a spoliation issue the firm briefed last year. The memo finds the same controlling case and one newer district-court opinion. One pinpoint is two pages off and one secondary source is cited for a point it only mentions in a footnote. She fixes both, notes the newer case for the partner, and keeps the prior brief as the benchmark.

Try this

Primary workflow:

Ask Protégé to draft one research memo on an issue your firm briefed last year. Have the associate check every citation's Shepard's status and pinpoint against the prior brief before the memo leaves the matter.

2. Agentic drafting to Word, Excel, PowerPoint, and PDF

What it is: Protégé Agentic Drafting uses purpose-built agents to draft contracts, motions, briefs, and deal documents grounded in LexisNexis content, firm templates, prior work, and matter materials, with outputs in Word, Excel, PowerPoint, and PDF (7 May 2026). The August update adds document-aware editing with track changes and suggested next steps, and search across a connected document management system.

Formatted output looks finished. The risk is that a polished draft gets less scrutiny than a rough one.

Concrete example (senior associate) A senior associate drafts a motion to compel from a firm template and the matter file. The draft is well formatted and follows the template. One meet-and-confer date is wrong because the agent read a draft letter rather than the sent one. She checks every date and fact against the file before the partner sees it.

Try this

Check facts in polished drafts:

Draft one motion your firm already filed from the same template. Have the senior associate compare every date, fact, and citation with the filed version before using agentic drafting on a live motion.

3. Shepard's citation checks

What it is: Shepard's citations sit under Lexis+ with Protégé answers and drafts. Since 7 May 2026, Shepard's Verify Trust Markers scan AI-generated and attorney-drafted text for legal citations, check them against LexisNexis sources, and flag citations that cannot be verified as existing; LexisNexis describes this as confirming that cited authority exists and is retrievable. Its product page goes further and says Shepard's Verify flags unsupported statements.

LawSites' 7 May 2026 report noted that an existence check is not the same as confirming a case supports the point it is cited for. Treat a clean marker as “this case is real”, then read it.

Concrete example (associate at a 25-lawyer litigation firm) An associate runs Shepard's Verify on a draft memo. It catches one citation with a wrong reporter volume. All other markers are clean, but one cited case was reversed in part on the exact issue; the Shepard's report shows it, the marker alone does not tell the story. She reads every cited case and adds the reversal to the memo.

Try this

Exists is not enough:

Run Shepard's Verify on one memo your team already cite-checked by hand. Have the associate compare every flag and every clean marker with the manual cite-check before relying on the markers.

4. Workrooms and customer-held encryption keys

What it is: Protégé Workrooms are private, permission-aware spaces where firms, clients, and co-counsel share documents, drafts, analysis, and AI workflows only as authorized, with dual approvals, least-privilege access, role-based permissions, and audit trails (7 May 2026). Protégé BYOK lets enterprise and government customers hold their own encryption keys in AWS KMS, Azure Key Vault, Google Cloud KMS, or HashiCorp Vault and revoke access; LexisNexis says it is already in use at AmLaw 100 firms.

Shared AI spaces across organizations raise privilege and confidentiality questions. Who can see a draft, and who can run AI on a co-counsel's documents, should be set before the first invitation goes out.

Concrete example (knowledge-management lead) A KM lead sets up a Workroom with co-counsel on a closed joint-defense matter. The test shows co-counsel can see a privileged internal draft because it sat in a shared folder. She moves the draft, requires dual approval for new shares, and writes a one-page sharing rule the responsible partner signs.

Try this

Permissions first:

Open one Workroom with co-counsel on a closed matter. Have the responsible partner check who can see each document and AI output before inviting anyone on a live matter.

5. Vault document analysis

What it is: Protégé Vault stores matter documents for AI tasks such as timelines, citation finding, counterarguments, and drafting, with outputs linking back to exact passages, spreadsheet rows, images, or timestamps (7 May 2026). LexisNexis pages give different limits: the May release says up to 100,000 documents per Vault, an August 2026 product post says capacity is expanding to 10,000 documents with files up to 500 MB and audio, video, and images (expected for U.S. customers in fall 2026), and the product FAQ still says up to 50 Vaults of 1–500 documents.

Ask your rep which limit is on your contract. The FAQ also says single uploads of up to 10 documents are purged at the end of a session and Vault results are kept 90 days.

Concrete example (litigation paralegal) A paralegal loads a closed matter's deposition exhibits into a Vault and asks for a timeline. Most entries link to the right page. Two dates come from an unsigned draft agreement in the set. She removes drafts from the Vault and asks the attorney to approve the source set before the next run.

Try this

Check the source set:

Build one Vault timeline from a closed matter. Have the paralegal click through every entry to its source and the handling attorney approve the source set before using Vault on a live matter.

6. Skills and Claude legal plugins

What it is: Skills are reusable instructions for tasks such as contract comparison, complaint analysis, research synthesis, checklists, due diligence, and playbook review (7 May 2026). LexisNexis put Anthropic's legal plugin into Protégé in a commercial preview on 24 Feb 2026 and integrated Anthropic's suite of 12 Claude legal plugins — including Product Counsel, Employment Counsel, AI Governance Counsel, and Litigation Associate — on 13 May 2026. The August update says customer-created skills and workflows are coming soon.

A skill encodes someone's way of working. A firm skill should have a named owner who reviews its output, just like a template.

Concrete example (practice-group partner) A partner tries a Litigation Associate skill for a complaint analysis on a closed case. It spots the same claims as the firm's memo but misses a notice requirement under a state statute. The partner adds a notice check to the firm's own checklist and names an owner for any firm skill built from it.

Try this

Name a skill owner:

Run one Claude legal skill on a closed matter. Have the practice-group partner compare its output with the firm's memo and name who owns that skill before anyone uses it on a live matter.

Pricing context (October 2026)

LexisNexis does not print a Lexis+ with Protégé price. Its product FAQ says pricing is customized by organization size, capabilities, and content scope and points to 1-888-AT-LEXIS; the online store for law firms FAQ says to call 888-AT-LEXIS to learn more about Lexis+ with Protégé (both checked 20 Oct 2026). We treat it as quote-only and do not print third-party per-attorney figures.

LexisNexis offers a two-day free trial on the product page. When you ask for a quote, ask which content set (jurisdictions, Practical Guidance) and which pieces — Workrooms, Vault limits, BYOK, Claude plugins — are included, and whether the fall 2026 Vault and Workroom updates are live in your account.

Uncertainty: We did not test Lexis+ with Protégé. Features and dates come from LexisNexis press releases, product pages, and blog posts read on 20 Oct 2026; several items were described as coming soon or expected in fall 2026, and Vault limits differ between LexisNexis pages. Document counts, ROI studies, and “most comprehensive” language are LexisNexis claims.

When it’s useful (honest take)

Useful when Lexis is your research platform of record, associates already Shepardize as a habit, and a named attorney will read every authority and check every fact in a polished draft before it leaves the matter.

Less useful when your firm runs on Westlaw (CoCounsel Legal), your main need is practice management (Clio AI), or you are a PI firm that mainly needs demands and chronologies (EvenUp — which now offers Protégé inside its Companion assistant for mutual customers — or Filevine AI). Large firms also compare it with Harvey AI, which offers Ask LexisNexis.

Related options (educational)

FAQ

Is Lexis+ AI the same as Lexis+ with Protégé?

LexisNexis renamed Lexis+ AI to Lexis+ with Protégé in February 2026 (general availability 24 Feb 2026). Its product FAQ says former Lexis+ AI users have access to the new platform.

How much does Lexis+ with Protégé cost in October 2026?

LexisNexis prints no price. Its product page says pricing is customized and points to 1-888-AT-LEXIS, and the law-firm store FAQ says to call for Lexis+ with Protégé. A two-day free trial is offered.

Does Shepard's Verify prove a citation supports my argument?

No. LexisNexis describes it as confirming cited authority exists and is retrievable, and it flags citations it cannot verify. A licensed attorney still reads the case and owns every citation in a filing.

What are the Claude legal skills in Protégé?

LexisNexis integrated Anthropic's legal plugin in a commercial preview on 24 Feb 2026 and a suite of 12 Claude legal plugins on 13 May 2026, covering areas such as product, employment, AI governance, and litigation work.

Who should check Lexis+ with Protégé output?

The associate for every authority's Shepard's status and pinpoint; the senior associate for facts and dates in drafts; the responsible partner for Workroom sharing; the handling attorney for the Vault source set.

Optional next step (if seats overlap)

If you pay for Lexis+ with Protégé, a second research platform, and general AI seats that all touch the same memo, that overlap is worth mapping. Map the overlap with the $19 AI toolkit audit or the $9 AI toolkit spreadsheet. Email Audit $19 or Toolkit spreadsheet $9.

Checkout is live via Stripe. After you pay, we email delivery (spreadsheet) or intake next steps (audit).

Related: CoCounsel Legal · EvenUp · Clio AI · Harvey AI · Filevine AI · Claude for legal ops · Gemini for legal ops · Spellbook (contract redlines) · Ironclad (in-house CLM) · Microsoft 365 Copilot · Claude · Three checks before another AI seat · ops hub