OpenAI launches Astra for Law, a GPT-6 configuration for legal research
SiliconANGLE Duncan Riley ● Covered by 2 sources
OpenAI wrapped GPT-6 Astra in a legal research setup called Astra for Law. It’s built for law firms, and its pitch is trust: better authority checks, tighter controls.
Based on reporting by SiliconANGLE, Duncan Riley — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI has turned GPT-6 Astra into something aimed squarely at lawyers. The new package, Astra for Law, is not a fresh model. It is GPT-6 Astra layered with a legal search index and instructions for legal analysis, then offered to law firms and the software vendors that serve them.
The search layer is broad. OpenAI says it reaches U.S. case law, statutes, regulations, court rules and administrative decisions across more than 230 million URLs, with new sources added daily. The nonprofit Free Law Project’s CourtListener database provides the case law. That matters because legal search is not just about finding an answer; it is about finding the right authority and the exact passage that supports it.
On that kind of task, OpenAI says Astra for Law beat GPT-6 Astra with web search alone on the same private validation questions from Vals AI’s Legal Research Bench. The legal version got 54% of 200 questions correct, while web search alone managed 38.7%, with both systems pushed to their highest reasoning effort. On case-law questions, it found 24% more reference cases. On another audited set, it pulled up to 54% more of the target passages from the correct opinions.
Access starts small. Selected firms can use it through a Trusted Access program in ChatGPT and Codex, where it appears in the model picker as GPT-6 Astra Law. The API version, gpt-6-astra-law, is coming later, but OpenAI gave no date or price. Harvey Technologies Inc. and Legora AB are already named as API customers.
OpenAI is also wrapping the launch in the kind of controls law firms care about. Eligible firms get zero data retention on the API, and ChatGPT Enterprise usage is excluded from human review by default. OpenAI is working with Latham & Watkins on governance design for permissions, ethical walls and client instructions. Alongside that, 26 partner-built plugins connect ChatGPT to tools including Relativity, Clio, iManage and DeepJudge, while Thomson Reuters Corp. is bringing HighQ matter context into ChatGPT and previewing a connector for CoCounsel Legal. Another nine plugins from outside lawyers and legal engineers add 47 custom skills. ChatGPT for Word also reached general availability the same day.
OpenAI says this is just the start of a longer push into law. That sounds right, and also very OpenAI: ship the wrapper, call it a platform, and let the firms pay to sort out what “trusted intelligence” means when the redlines start flying.
My take — AI-written commentary, not fact-checked reporting
This is the correct move: not a grand new model, just a model boxed up for a job that hates sloppy answers. Legal AI keeps drifting toward the same truth — the winning product is governance with a model attached, not the other way around. And yes, the trust pitch is doing a lot of work here; that’s usually because the work itself is hard.
Read more about this at: SiliconANGLE
