Startups & Funding

Anthropic expands Claude for Legal with new plug-ins and MCP connectors

New tools automate document search, drafting, and deposition prep for law firms.

Deep Dive

Anthropic announced Tuesday a host of new chatbot features for law firms, expanding its Claude for Legal platform with specialized plug-ins and model context protocol (MCP) connectors. The plug-ins bundle automated functions for commercial, privacy, corporate, employment, product, and AI governance law, handling clerical tasks like document search/review, case law research, deposition prep, and document drafting. The MCP connectors integrate Claude directly with commonly used legal software—Docusign for document management, Box for file search, and Thomson Reuters (including Westlaw) for legal research. These tools are available to all paying Claude customers, building on earlier legal plug-ins launched in February.

The launch comes amid fierce competition in the AI legal services space. In March, AI law startup Harvey raised $200 million at an $11 billion valuation, using agentic AI to automate legal workflows. Last month, rival Legora secured a $600 million Series D and launched a campaign with Jude Law, offering similar automated solutions. These developments follow a pattern of AI misuse in courts—dozens of lawyers have been caught using AI to generate error-ridden documents, and California fined an attorney for ChatGPT-drafted appeals with fake quotes. 'The legal sector is facing mounting pressure to adopt AI,' an Anthropic spokesperson said, 'and the firms that move are pulling ahead fast.'

Key Points
  • Anthropic launched legal-specific plug-ins for commercial, privacy, corporate, and employment law, automating document drafting and review.
  • New MCP connectors integrate Claude with Docusign, Box, and Thomson Reuters (Westlaw) for direct data access.
  • Competitors Harvey (raised $200M at $11B valuation) and Legora ($600M Series D) are racing to automate legal workflows.

Why It Matters

Automating clerical legal tasks could reduce costs but risks AI errors already causing court sanctions.