OpenAI Launches Astra for Law With 230 Million-URL Legal Search Index

OpenAI has launched Astra for Law, a legal-tailored configuration of its most powerful model, GPT-6 Astra, paired with a new search index spanning more than 230 million URLs of U.S. case law, statutes, regulations, court rules, and administrative decisions. The company says the product is a foundation for law firms and legal technology companies to build AI products and workflows around their own expertise.

The stakes are significant. Legal research and drafting are among the highest-value knowledge-work tasks AI vendors are competing to automate, and OpenAI is positioning Astra for Law as infrastructure for the entire legal technology ecosystem rather than a standalone product. API customers including Harvey and Legora will be able to build on Astra for Law, bringing the model's capabilities into their own products and workflows. OpenAI says it will extend these legal capabilities to its frontier models as they advance.

A 40% relative benchmark jump

OpenAI's central claim rests on hard numbers. The company tested Astra for Law's complete setup on 200 U.S. legal research questions drawn from the private validation set of Vals AI's Legal Research Bench, which measures how well a model finds relevant sources and passages and how well its answers meet evaluation criteria.

At the highest reasoning effort for both systems, Astra for Law passed the evaluation's overall correctness check on 54.0% of questions, compared with 38.7% for GPT-6 Astra using web search alone — a 40% relative improvement. On case-law-focused questions, Astra for Law found 24% more reference cases than the base model with web search. On the audited set of target passages, it retrieved up to 54% more relevant passages from the correct court opinions at the same reasoning effort. OpenAI also says Astra for Law produces more comprehensive answers.

The search index that drives these gains has a notable open-data backbone. OpenAI's work with Free Law Project, the nonprofit behind CourtListener, brings a case-law collection covering more than 99.9% of published U.S. precedential case law into the research experience, with sources added daily. The index complements the licensed content and specialist products firms already rely on from providers such as Thomson Reuters.

Head-to-head with Anthropic's frontier model

OpenAI published a comparison that puts a competitor on the back foot. When prompted to identify good law with similar fact patterns — a scenario involving a five-year private-label manufacturing agreement, a $2 million annual purchase commitment against a claimed $8 million historical average, and a potential misrepresentation claim — Astra for Law returned two closely matching precedents. Anthropic's Claude Fable 5.1, given the same prompt, returned a holding that had been reversed on appeal in the litigation example and reported finding no such case in the transactional example.

The example illustrates what OpenAI says is the product's differentiator beyond retrieval: custom instructions for legal analysis and writing that guide the model in applying research to a client's facts, developing arguments or deal terms, and identifying weaknesses and uncertainty. That can mean distinguishing a court's holding from its other observations, addressing cases that weaken an argument, or explaining how a contract exception shifts risk between parties.

Trusted Access, Zero Data Retention, and ethical walls

Confidentiality governs everything in legal practice, and OpenAI has built the launch around that constraint. Astra for Law will be initially offered to selected law firms through Trusted Access in ChatGPT and Codex, with API availability coming soon. It appears in the model picker as "GPT-6 Astra Law" and in the API as gpt-6-astra-law.

For eligible firms, the offering includes Zero Data Retention on the API, and usage of ChatGPT Enterprise is excluded from human review by default. OpenAI is also working with Latham & Watkins, which it describes as a leader in AI governance, to design for information permissions, ethical walls, client instructions, and firm oversight.

Big Law builds its own tools

The launch showcases what OpenAI's forward-deployed engineers have built with selected firms on ChatGPT Enterprise, with custom interfaces and integrations into proprietary data.

Sullivan & Cromwell built an agreement analyzer that brings the firm's negotiating playbooks and selected precedents into the review of a new deal, helping lawyers spot risks that emerge when provisions are read together, then turning findings into proposed redlines and draft client advice. Ropes & Gray built a deal diligence system modeled on how its lawyers work through a data room, tracing findings back to source documents and pinpointing questions that could affect an acquisition, such as whether key customer contracts require notice or consent. Cooley built GO Public, a tool that carries the firm's capital markets expertise into IPO preparation, from drafting the filing to identifying risks that deserve management's attention — and propagating changes across the filing when the deal terms move.

The pattern matters for the market: these are firm-specific systems built on OpenAI's platform, not generic assistants. As the company puts it, firms can turn their own precedents, methods, and judgment into AI tools and workflows built to their standards, with permitted sources and review processes defined for the work.

26 plugins, 47 skills, and Word integration

OpenAI is launching 26 partner-built plugins that connect ChatGPT to specialist legal tools. With iManage, a lawyer can draft a negotiation brief in ChatGPT and save it to the matter file. Intapp can surface activities that may need a time entry for review. DeepJudge can bring prior deals into a comparison. Thomson Reuters is bringing HighQ matter context into ChatGPT and previewing a forthcoming CoCounsel Legal connector.

The launch also includes 9 community plugins from lawyers and legal engineers at LegalQuants, LECG, and Skills.law, along with 47 custom skills that practitioners can adapt, extend, or draw inspiration from. ChatGPT for Word becomes generally available today, letting lawyers proofread, get suggested edits, and flag formatting issues inside the drafting tool they already use.

OpenAI frames its approach as open and composable: firms can use specialist products, bring their own tools and knowledge, and adapt community-built skills. "Building on OpenAI should mean getting more from the ecosystem, not replacing it," the company says.

Long-term bet, with Wachtell

OpenAI states plainly that it is investing in law for the long term, committing to keep advancing Astra for Law's model, settings, tools, and instructions together, guided by rigorous evaluations and feedback from lawyers and legal technology partners. Its work with Wachtell, Lipton, Rosen & Katz brings the firm's litigation and corporate expertise together with OpenAI's frontier research and engineering to explore how AI can support the work involved in providing sophisticated legal judgment.

The competitive picture is now explicit: OpenAI is not just selling a model to lawyers but courting Harvey, Legora, Thomson Reuters, and the largest U.S. firms as platform partners — a play to become the underlying intelligence layer for legal work while leaving firms and vendors to own the workflows. Whether the 54.0% benchmark figure holds up under independent scrutiny, and whether ethical walls and Zero Data Retention satisfy malpractice insurers and regulators, will determine how fast Big Law moves from pilot to production.