Legal Document Analysis

Legal teams use AI to search, compare, review, and classify contracts and other legal documents.

Business impact

• Search speed — Microsoft reports UBS can complete complex queries within minutes rather than hours, and Mitsubishi reports improved search speed.
• Search precision — Mitsubishi reports improved search precision after loading contract data into MNTSQ for Enterprise.
• Time savings — Vodafone’s legal team is reported to have time savings from Microsoft 365 Copilot, and UBS reports faster query completion.

Data requirements

• Contracts (text) — The systems use accumulated contract data and contract repositories.
• Legal documents (text) — ASUS and UBS use legal documents and clauses as the input corpus for analysis and search.
• Clause-level text (text) — Mitsubishi and UBS both refer to clause or paragraph-level retrieval and analysis.

AI methods and techniques

• Natural-language retrieval — Used to let legal staff search documents with ordinary-language questions.
• Semantic search — Used to surface relevant clauses and paragraphs even when the exact wording differs.
• Contract classification — Used at Mitsubishi to automatically classify contract types.
• Clause-level analysis — Used to analyze contract clauses and support comparison and review.
• Document summarization — Used in Vodafone’s trial to summarize documents.
• Text generation — Used in JPMorgan Chase for generating contracts and in other review and drafting work.

AI models and model families

LLM Suite, Azure OpenAI Service, Microsoft 365 Copilot

Industries

Real-world evidence

7 documented case studies on record.

Companies using this: ASUS, JPMorgan Chase, Mitsubishi, Socure, UBS, Vodafone, Weaviate.