In a major announcement involving players from across the legal tech ecosystem, OpenAI today (17 September) unveils Astra for Law – it’s most powerful large language model to date configured into a new foundation for law firms and legal technology companies to build on.
Astra for Law combines GPT-6 Astra with settings, tools, and context tailored for professional legal work. It also launches with 26 partner plugins from providers including Thomson Reuters, Intapp, Harvey, Legora, DeepJudge and iManage.
Advertisement
One of the new tools is a new legal search index, through which users can search for case law, statutes, regulations and court rules. OpenAI has long collaborated with the Free Law Project and this latest move makes its caselaw collection available through the platform.
OpenAI has tested Astra for Law on 200 U.S. legal research questions from the private validation set of Vals AI’s Legal Research Bench. It shows a 40% improvement in correctness compared with GPT-6 Astra using web search alone. On case-law-focused questions, Astra for Law found 24% more reference cases than GPT-6 Astra using web search alone at the highest reasoning effort.
Advertisement
There is no benchmark available comparing it with leading legal research providers, and OpenAI notably suggests that the new functionality “complements the licensed content and specialist products firms rely on from providers such as Thomson Reuters.”
At early adopter US multinational law firm Sullivan & Cromwell, partner John Savva told Legal IT Insider: “Astra for Legal presents a significant opportunity to consolidate two workflows that have, until now, remained largely separate: researching the law and analyzing legal sources. To date, lawyers have generally had to complete their research first, then use AI to help with analysis, because no model has had the capacity to serve as the sole source of legal authority. Astra for Legal brings us much closer to the day when legal research and analysis can be conducted under one hood, making the legal research and analysis process more effective and efficient.”
Oz Benamram, chief AI officer at Pillsbury and founder of SKILLS.law, told us: “This is an important development. We are happy that OpenAI is taking the legal sector seriously and Pillsbury is excited to work with OpenAI on building tools that help our lawyers and clients use the platform.”
Plugins and APIs
The 26 partner plugins cover the practice and business of law. With iManage, for example, 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.
API customers including Harvey and Legora will be able to build on Astra for Law, bringing this intelligence into their own products and workflows.
Speaking to a group of journalists ahead of the launch, Jason Boehmig, the head of OpenAI’s legal vertical and former founder of Ironclad (seen below at the announcement today), said: “I wanted to make sure we led with an ecosystem-forward approach,” adding, “We want to be the home for builders, and arm people with the tools and capability to do that.”
Joel Hron, chief technology officer of Thomson Reuters, said that legal professionals will need to continue to access trusted intelligence and the governance required for high stakes work, such as that found in CoCounsel and HighQ, observing: “Our work with OpenAI helps make these capabilities available in the environments customers choose, while preserving the accuracy, confidentiality, and accountability they depend on.”
OpenAI is working with Latham & Watkins to design for information permissions, ethical walls, client instructions, and firm oversight. Michael Rubin, chair of Latham’s AI strategy committee, said: “As AI becomes more capable, so too does the ability to deploy it in environments that demand rigorous governance, oversight, and accountability. This collaboration builds on Latham’s broader investments in AI development, governance, and infrastructure across the firm, which underpin our enterprise-wide strategy for responsible AI.”
OpenAI says its early trusted access program includes includes Zero Data Retention, and usage of ChatGPT Enterprise is excluded from human review. There is a thirty-page agreement setting out exactly what that means.
“We were grateful to preview early versions of Astra for Law, which were built with legal use-cases in mind,” said Savva. “Across both litigation and transactional matters, the models demonstrated impressive research depth and sensitivity to authority. Even at this early stage, they felt like a significant step toward legal-focused AI that is carefully grounded in research that is both current and comprehensive.”
ChatGPT Enterprise custom interfaces
OpenAI has shared live examples of how firms can turn their own precedents, methods, and judgment into AI tools and workflows built to their standards.
Behind the scenes, OpenAI’s engineers have been working with early adopter firms to adapt ChatGPT Enterprise with custom interfaces and integrations to proprietary data. Examples include:
● Sullivan & Cromwell has built an agreement analyzer that brings the firm’s negotiating playbooks and selected precedents into the review of a new deal;
● Ropes & Gray has built an M&A diligence system around how its lawyers work through a data room and decide what matters to the deal;
● Cooley has built GO Public to bring its capital markets expertise into how companies prepare to go public, from drafting the IPO filing to identifying the risks that deserve management’s attention.
“Our collaboration with OpenAI has allowed us to rethink how this work gets done – moving lawyers and management teams more quickly through intensive preparation and into questions that require judgment, market experience and strategic thinking,” said Dave Peinsipp, partner and co-chair of Cooley’s global capital markets group.
Savva told Legal IT Insider: “Astra for Legal makes it easier for firms to build tools that leverage their proprietary knowledge. With a base layer of legal knowledge built in, firms can focus on the expertise and approaches that make them distinctive.”
Community plugins
The launch includes nine community plugins from lawyers and legal engineers at LegalQuants, LECG, and Skills.law, with 47 custom skills that practitioners can adapt, extend, or draw inspiration from to do their work in ChatGPT. Community plugins give the people closest to the work a way to keep setting the standard for what these skills can do.
OpenAI also said it is making ChatGPT for Word generally available today, so lawyers can proofread, get suggested edits, and flag formatting issues in the tool they already rely on for drafting.
It’s interesting to note that OpenAI launched ChatGPT for Financial Services just over a week ago and speaking to journalists ahead of today’s launch, Sherwin Wu, a member of technical staff at OpenAI (seen below at the announcement today), said that the teams will be working together on certain features.
Key takeaways
OpenAI is not the first to announce a legal model configuration – Claude for Legal launched in March. But the leap in capabilities of Astra in getting a piece of work done from start to finish, combined with the extent to which OpenAI is investing in configuring the model and building an ecosystem around it, will prove extremely powerful.
The previous generation of legal GenAI largely asked: ‘How good is the output?’ The reality is that Astra pushes the question towards: ‘How much of the workflow can AI complete before a lawyer needs to intervene?’