Products·2 min read
By BitsMindsSource: Anthropic

Anthropic Pushes Claude Cowork Into Big Law With 20+ Integrations and an M365 Agent

12 role-specific plugins, a cross-app Microsoft 365 agent, and live deployments at Freshfields and Quinn Emanuel — Anthropic is making legal its biggest Cowork beachhead.

Anthropic Pushes Claude Cowork Into Big Law With 20+ Integrations and an M365 Agent
Share:

Anthropic is going hard at the legal market, rolling out more than 20 new Claude Cowork integrations and 12 role-specific AI plugins built around the way law firms and in-house teams actually work — contract review, redlining, M&A due diligence, employment handbook drafting, and litigation matter management. The push, unveiled around an Anthropic webinar that drew more than 20,000 legal professionals on May 15, is the company's largest vertical-specific Cowork expansion to date and lands legal at the top of the power-user leaderboard inside its agentic platform.

Watch: Claude Cowork for legal teams (Anthropic on YouTube)

The headline integration is a cross-app Microsoft 365 agent. Claude now embeds across Microsoft 365 — Word, Outlook, Excel, and PowerPoint — as a single context-carrying agent, so a lawyer can move from drafting a brief to summarizing a discovery email thread to building a deal model without losing state. That matters because the workflow context, not raw model quality, has historically been the bottleneck for AI in legal: every app handoff was a fresh prompt and a fresh chance to hallucinate.

Anthropic is also leaning hard on grounding. Claude Cowork now connects through plugins to Westlaw, CourtListener, and iManage repositories, anchoring outputs in verified primary sources rather than free-form generation. "In litigation, an authoritative-sounding hallucination is worse than no answer," Eve's CEO told Fortune — a line that captures both the technical bet and the liability calculus driving the design.

The early adopter list is unusually heavyweight: Freshfields, Quinn Emanuel Urquhart & Sullivan, Holland & Knight, and Crosby Legal are all running Claude on active matters. Quinn Emanuel partner Christopher Kercher, describing how the firm built its litigation platform, said the unlock was philosophical: "The breakthrough was treating Claude like a member of the case team." Anthropic is also citing a 90.9% score for Claude Opus 4.7 on Harvey's BigLaw Bench, the closest thing the industry has to a frontier-model legal benchmark.

The strategic read is that Anthropic is doing for legal what it just did for finance with the PwC Office of the CFO deal: pick a high-margin, regulated, document-heavy function, build the connectors and plugins that match the workflow, and stake out the category before OpenAI or Microsoft's own Copilot Legal Agent can lock it down. Combined with the Big Law adopter list and the M365 cross-app agent, Cowork for legal is positioned as one of the most concrete agentic deployments in the enterprise market right now.

More on Claude

Evergreen coverage we keep current — start here.

Want AI news before everyone else?

The morning's most important AI stories, straight to your inbox. No fluff.

Related Articles

Claude Code: a little room to wrap up On a warm cream workbench, a charcoal terminal carries the terracotta Claude emblem and a full usage meter marked 100%. The meter has a small striped continuation beyond its limit. A jointed metal arm holds a large terracotta closing curly brace above a sheet of code, over a similarly striped edge. The terminal and desk label both read WRAPPING UP, and the desk label specifies CURRENT RESPONSE. The image is an original metaphor for continuing a response briefly toward a reasonable stopping point after reaching the five-hour usage limit. It does not depict a guaranteed completed task or a specific amount of extra usage. Original BitsMinds editorial SVG, 6 October 2026, for claude-code-wrap-up-allowance-usage-limit. Context: the article's local publication text and https://support.claude.com/en/articles/17040437-claude-code-wrap-up-allowance. The allowance is small, capped and for an already running response; it counts toward the existing weekly cap. This is symbolic artwork, not an exact product screenshot. The striped segment is not to scale. Vector-only artwork with the project Claude symbol. CLAUDE CODE USAGE LIMIT 100% WRAPPING UP WRAPPING UP CURRENT RESPONSE BITSMINDS.COM
Products

Claude Code Now Gets a Grace Period at the Usage Limit

A hidden watermark in ChatGPT text A printed page of text lies on an EU-blue desk ringed by faint gold stars. A round detector lens over the page shows the same lines as a coloured grain of marked words, with the OpenAI logo surfacing inside the lens. TEXTGRAIN EU · AI ACT BITSMINDS.COM
Products

OpenAI Watermarks ChatGPT Text in the EU With textGrain

Gemini's free plan, locked to Flash-Lite A heavy graphite padlock with the official Gemini sparkle inlaid in its body sits on a pale field lit by soft Gemini-coloured glows. On the left, three model chips: Flash-Lite is open with a green tick, Flash and Pro are greyed out behind small locks. A paper tag tied to the shackle reads Free plan, from 9 October. FREE PLAN from 9 October Flash-Lite Flash Pro BITSMINDS.COM
Products

Free Gemini Users Drop to Flash-Lite Only From 9 October