Westlaw AI

Westlaw is the gold standard for American legal research, and Westlaw AI (powered by Thomson Reuters' CoCounsel and generative AI integration) brings conversational AI to the most comprehensive legal database available.

## The Research Platform

Westlaw's database includes: - Every federal and state court decision - All federal and state statutes and regulations - Secondary sources (treatises, law reviews, practice guides) - International and foreign law in 50+ jurisdictions

When the AI answers a legal question, it grounds responses in this authoritative source material — not internet-scraped summaries.

## AI Capabilities

**Legal question answering** — ask in plain English: "What's the standard for preliminary injunctions in the Ninth Circuit?" The AI synthesizes relevant case law and provides a reasoned answer with citations.

**Case summarization** — upload or identify a case; AI generates executive summaries, extracts key holdings, and identifies relevant precedents.

**Statute analysis** — point the AI at a statutory section; it explains plain-language meaning, identifies judicial interpretations, and flags circuit splits.

**Litigation strategy** — AI analyzes how courts have ruled on specific arguments to help attorneys assess case strength before filing.

**Document review** — identify relevant clauses, flag issues, and extract key terms from contracts and briefs.

## CoCounsel Integration

Thomson Reuters integrates Microsoft Azure OpenAI models under the "CoCounsel" brand. CoCounsel is designed for attorney use with hallucination safeguards: every AI response is sourced to specific Westlaw documents, with clickable citations that let attorneys verify the underlying authority.

## Audience

Used by 90%+ of Am Law 100 firms, federal government legal departments, and law schools. Subscription pricing is enterprise-scale — individual attorneys typically access through firm subscriptions.

## Compliance

Thomson Reuters maintains attorney-client privilege protections in its data handling and has published specific assurances about not training AI on user queries.