Editorial illustration for OpenAI Targets Legal Market With Astra's 230 Million Case Law Index
OpenAI Launches Astra for Law With 230M Case Index
OpenAI Targets Legal Market With Astra's 230 Million Case Law Index
OpenAI has built a legal research tool on top of its GPT-6 Astra model, and it wants law firms to pay attention. Astra for Law comes bundled with a search index covering more than 230 million URLs of US case law, statutes, and regulations, drawing on data from the Free Law Project, which claims coverage of over 99.9 percent of published US precedents. The model itself doesn't just search that index; it also carries instructions built for legal analysis and legal writing, distinguishing it from a general-purpose chatbot pointed at a law library.
OpenAI is pitching this at multiple layers of the legal market at once. API access means companies like Harvey and Legora can build their own products on top of Astra for Law. Firms that want to use it directly get something called Trusted Access, which includes privacy controls such as zero data retention. There are also 26 plugins going out for software already sitting on lawyers' desktops, including Relativity and Clio.
The timing isn't accidental. Anthropic has been pushing into the same legal work market, and OpenAI's own benchmark numbers, run internally, are about to make the case for why its approach should win out.
The index searches US case law, statutes, and regulations across more than 230 million URLs. OpenAI draws on data from the Free Law Project, which says it covers over 99.9 percent of published US precedents. In a test OpenAI ran itself using Vals AI's Legal Research Bench, Astra for Law passed 54 percent of the 200 questions, compared to 38.7 percent for GPT-6 Astra with plain web search.
Why this matters
A 54 percent pass rate on Vals AI's Legal Research Bench is not a number OpenAI should want printed next to a product built for lawyers, and yet here it is, in their own benchmark. That's the detail worth sitting with. Coverage of case law isn't the hard part anymore, the Free Law Project has basically solved that with 99.9 percent of published precedent indexed across 230 million URLs.
The hard part is what a model does with all that material once it's in context, and a coin-flip-adjacent score on a research task suggests Astra for Law is closer to a capable first draft tool than something you'd trust unsupervised on a brief. For founders building in legal tech, this is both a warning and an opening: the retrieval layer is getting commoditized fast, so the differentiation has to come from verification, citation-checking, and workflow fit, not from who has the bigger index. For anyone evaluating vendor claims generally, the lesson is simple: read the benchmark number before the marketing copy.
Related Reading
Browse all OpenAI news →
Policy & Regulation Akamai Proposes SAFE Cybersecurity Guidelines for AI Threats
Policy & Regulation After ditching AI fitness apps and a Fitbit, I return to Peloton classes
Policy & Regulation Sony Music, Warner Chappell Sue Anthropic Over "Blatant Theft
Related Topic OpenAI researcher quits, citing distrust over ad‑driven engagement metrics
LLMs & Generative AI
Related Topic OpenAI launches GPT-Image 1.5 with precise editing for enterprise visuals
LLMs & Generative AI
From Archives AI Models Solve Millennium Math Problems, Experts Warn of Urgent Risks
From Archives Bipartisan Frontier Act Proposes Independent AI Auditors
From Archives Microsoft and OpenAI Workers Warn AI Is Harming the Web
Research & Benchmarks
From Archives OpenAI Developer: AI Agent "Swarms" Waste Tokens, Create "Coordination Tax
AI Tools & Apps
Common Questions Answered
What is Astra for Law and how does it differ from standard GPT-6 Astra?
Astra for Law is OpenAI's specialized legal research tool built on the GPT-6 Astra model, designed specifically for law firms and legal professionals. Unlike the standard GPT-6 Astra, it comes bundled with a search index covering over 230 million URLs of US case law, statutes, and regulations, and includes built-in instructions optimized for legal analysis and legal writing tasks.
How comprehensive is the case law coverage in Astra for Law's index?
Astra for Law's index covers more than 99.9 percent of published US precedents, drawing on data from the Free Law Project across 230 million URLs. This extensive coverage includes US case law, statutes, and regulations, making it one of the most comprehensive legal databases available for AI-powered research.
What were the results of OpenAI's testing of Astra for Law on the Legal Research Bench?
In OpenAI's own testing using Vals AI's Legal Research Bench, Astra for Law passed 54 percent of 200 questions, compared to 38.7 percent for standard GPT-6 Astra using plain web search. This 15.3 percentage point improvement demonstrates the advantage of specialized legal training and dedicated case law indexing for legal research tasks.
What is the primary challenge that remains for legal AI models according to the article?
According to the article, comprehensive coverage of case law is no longer the hard part, as the Free Law Project has essentially solved that with 99.9 percent indexing across 230 million URLs. The real challenge is what the model does with all that material once it's in context, meaning the quality of legal analysis and reasoning is more critical than data availability.
Further Reading
- Astra For Law - OpenAI
- OpenAI launches Astra for Law with its own case law index - The Next Web
- OpenAI unveils Astra for Law - Legal Technology
- OpenAI Announces Astra for Law With GPT-6 Astra, Legal Search Index and Specialist Plugins - Gadgets 360
- OpenAI launches Astra for Law, says it will help lawyers with cases - India Today
