OpenAI unveils Astra for Law with legal search and firm controls
OpenAI has introduced Astra for Law, a legal AI foundation combining its GPT-6 Astra model with legal-specific settings, tools and context for law firms and legal technology companies. The offering includes a legal search index covering more than 230 million URLs and will initially be available to selected firms through Trusted Access in ChatGPT and Codex.
In the model picker, the product will appear as GPT-6 Astra Law; its planned API name is gpt-6-astra-law. API customers including Harvey and Legora will be able to build products and workflows on the foundation as the API offering becomes available.
Legal research index improves benchmark results
The new search index is designed to find relevant authorities, passages and their legal relevance. It searches US case law, statutes, regulations, court rules and administrative decisions, with new sources added daily. OpenAI said the index complements licensed content and specialist products used by firms, including services from Thomson Reuters.
Its work with Free Law Project brings CourtListener’s case-law collection into the experience. OpenAI said that collection covers more than 99.9% of published US precedential case law.
OpenAI evaluated the full Astra for Law configuration on 200 US legal research questions from the private validation set of Vals AI’s Legal Research Bench. At the highest reasoning effort, Astra for Law passed the benchmark’s overall correctness check on 54.0% of questions, compared with 38.7% for GPT-6 Astra using web search alone. OpenAI described this as a 40% relative improvement.
For case-law-focused questions at the same reasoning setting, Astra for Law found 24% more reference cases. On an audited set of target passages, it retrieved up to 54% more relevant passages from the correct court opinions. The product also uses custom instructions for legal analysis and writing, including identifying uncertainty, distinguishing holdings from other judicial observations, and addressing authorities that weaken an argument.
Controls, integrations and firm-built workflows
OpenAI is pairing the product with governance measures for confidential client work. Eligible firms in the Trusted Access Program receive Zero Data Retention on the API, while ChatGPT Enterprise usage is excluded from human review by default. The company is working with Latham & Watkins on designs for information permissions, ethical walls, client instructions and firm oversight.
Firms can also configure ChatGPT Enterprise around proprietary data, interfaces and workflows. OpenAI cited examples from Sullivan & Cromwell, Ropes & Gray and Cooley, including agreement analysis, transaction diligence and IPO preparation tools that retain links back to source material for lawyer review.
The launch adds 26 partner-built plugins connecting ChatGPT to legal systems such as Relativity, Clio, iManage, Intapp, DeepJudge and Thomson Reuters HighQ. It also includes nine community plugins containing 47 custom skills, and makes ChatGPT for Word generally available for proofreading, suggested edits and formatting checks.
Practical implication for legal teams
Legal teams considering Astra for Law will need to define permitted sources, review processes and access controls before deploying it on matters. The reported research gains and integrations may support research and drafting workflows, but lawyers remain responsible for examining authorities, testing outputs against client facts and applying professional judgment.