OpenAI released a legal industry-focused version of its GPT-6 Astra model on Thursday, escalating the competition among AI labs to capture business from law firms and corporate legal departments. The new platform, called Astra for Law, combines the company's latest AI technology with an index of US case law, statutes, regulations and other legal materials, along with specialized instructions for legal analysis and writing.
The launch follows Google's expansion of its Gemini Enterprise offerings for legal professionals last month, and tools released by Anthropic since January that let lawyers use its Claude AI assistant. Legal AI vendors including Harvey and Legora will be able to build products on Astra for Law, and the platform will include integrations with software providers such as Relativity, Clio, Intapp and Thomson Reuters.
Thomson Reuters, which owns Reuters and the Westlaw legal research database, launched its own AI model called Thomson 1.0 last month. The company said that model is built on open-source technology and draws on decades of its own legal content. It is trained on the company's legal research materials and built for professional work.
How law firms will access the platform
OpenAI said selected law firms will gain access through a special program that includes protections for confidential client work. The company has worked with firms including Sullivan & Cromwell, Ropes & Gray, Cooley, Latham & Watkins and Wachtell Lipton on testing and developing legal AI applications.
The platform is designed to help law firms and legal software providers conduct research, draft advice and build custom applications for lawyers. Rather than offering a generic AI tool, OpenAI has tailored the system specifically for legal workflows, embedding it with the structured legal knowledge that attorneys rely on daily. This push into specialized AI for Legal work signals how major AI developers now see professional services as a primary growth market.
An intensifying race for the legal market
The legal industry has become a key battleground for AI companies. Law firms generate vast amounts of billable work around research, document review and drafting - tasks that large language models can accelerate. OpenAI's move embeds its technology directly into the legal software ecosystem through partnerships with established vendors, rather than asking firms to adopt a standalone tool.
Anthropic and Google have taken similar approaches, each positioning their AI assistants as secure platforms that can handle sensitive client information while improving efficiency. Thomson Reuters, with its deep library of proprietary legal content, is betting that domain-specific training data will give its model an edge over general-purpose alternatives.
Why this matters for legal professionals
Law firms evaluating AI tools now face a market with competing platforms from OpenAI, Google, Anthropic and Thomson Reuters - each integrated with different legal software providers. The decision is no longer about whether to adopt AI, but which ecosystem to build on. Firms testing Astra for Law will need to assess how well the specialized legal instructions and case law index perform against alternatives trained on proprietary legal databases. The integrations with existing practice management tools like Clio and Intapp may determine how easily the technology fits into current billing and document workflows.
Your membership also unlocks:
