OpenAI introduced Astra for Law on September 17, 2026 as a legal AI foundation for law firms and legal technology companies. The offering combines GPT-6 Astra with legal-specific settings, tools, context, a legal search index, and instructions for legal analysis and writing. API customers including Harvey and Legora will be able to build Astra for Law into their own products and workflows, while selected law firms will first get access through Trusted Access in ChatGPT and Codex.

Astra for Law can search U.S. case law, statutes, regulations, court rules, and administrative decisions across more than 230 million URLs, with sources added daily. OpenAI said its 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. The system also complements licensed content and specialist legal products from providers such as Thomson Reuters.

OpenAI tested Astra for Law on 200 U.S. legal research questions from the private validation set of Vals AI’s Legal Research Bench, which measures source finding, passage retrieval, and research-answer quality. 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 and retrieved up to 54% more relevant passages from correct court opinions when systems were compared at the same reasoning effort.

OpenAI said Astra for Law uses custom instructions to help with end-to-end legal workflows after research, including applying authorities to client facts, developing arguments or deal terms, and identifying weaknesses and uncertainty. The company said the model can distinguish a court’s holding from other observations, address cases that weaken an argument, and explain how a contract exception shifts risk between parties. The product will appear in the model picker as “GPT-6 Astra Law” and in the API as gpt-6-astra-law.

OpenAI also announced privacy and governance controls for eligible law firms, including Zero Data Retention on its API and default exclusion of ChatGPT Enterprise usage from human review. The company said it is working with Latham & Watkins on information permissions, ethical walls, client instructions, and firm oversight. Law firms can customize Astra for Law through 26 new ecosystem plugins that connect ChatGPT to specialist legal tools such as Relativity and Clio.

OpenAI said selected firms have already built legal workflows with its forward-deployed engineers. Ropes & Gray built a deal diligence system for data-room review, Cooley built GO Public for IPO filing preparation and risk identification, and another firm built an agreement analyzer using negotiating playbooks and selected precedents. The launch also includes 26 partner-built plugins, including tools from iManage, Intapp, DeepJudge, and Thomson Reuters, plus 9 community plugins from LegalQuants, LECG, and Skills.law with 47 custom skills. OpenAI also made ChatGPT for Word generally available so lawyers can proofread, receive suggested edits, and flag formatting issues in Microsoft Word.

OpenAI introduced Astra for Law on September 17, 2026 as a legal AI foundation for law firms and legal technology companies. The offering combines GPT-6 Astra with legal-specific settings, tools, context, a legal search index, and instructions for legal analysis and writing. API customers including Harvey and Legora will be able to build Astra for Law into their own products and workflows, while selected law firms will first get access through Trusted Access in ChatGPT and Codex.

Astra for Law can search U.S. case law, statutes, regulations, court rules, and administrative decisions across more than 230 million URLs, with sources added daily. OpenAI said its 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. The system also complements licensed content and specialist legal products from providers such as Thomson Reuters.

OpenAI tested Astra for Law on 200 U.S. legal research questions from the private validation set of Vals AI’s Legal Research Bench, which measures source finding, passage retrieval, and research-answer quality. 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 and retrieved up to 54% more relevant passages from correct court opinions when systems were compared at the same reasoning effort.

OpenAI said Astra for Law uses custom instructions to help with end-to-end legal workflows after research, including applying authorities to client facts, developing arguments or deal terms, and identifying weaknesses and uncertainty. The company said the model can distinguish a court’s holding from other observations, address cases that weaken an argument, and explain how a contract exception shifts risk between parties. The product will appear in the model picker as “GPT-6 Astra Law” and in the API as gpt-6-astra-law.

OpenAI also announced privacy and governance controls for eligible law firms, including Zero Data Retention on its API and default exclusion of ChatGPT Enterprise usage from human review. The company said it is working with Latham & Watkins on information permissions, ethical walls, client instructions, and firm oversight. Law firms can customize Astra for Law through 26 new ecosystem plugins that connect ChatGPT to specialist legal tools such as Relativity and Clio.

OpenAI said selected firms have already built legal workflows with its forward-deployed engineers. Ropes & Gray built a deal diligence system for data-room review, Cooley built GO Public for IPO filing preparation and risk identification, and another firm built an agreement analyzer using negotiating playbooks and selected precedents. The launch also includes 26 partner-built plugins, including tools from iManage, Intapp, DeepJudge, and Thomson Reuters, plus 9 community plugins from LegalQuants, LECG, and Skills.law with 47 custom skills. OpenAI also made ChatGPT for Word generally available so lawyers can proofread, receive suggested edits, and flag formatting issues in Microsoft Word.