Law firm Ropes & Gray is working with OpenAI to build a proprietary AI system intended to speed up and standardize the data-heavy initial phase of transactional diligence, when buyers verify sellers’ financial and legal claims before acquisitions.
Why the tool matters
Transactional diligence typically requires large-scale document review under tight deadlines to uncover risks and liabilities. That work often involves an entire team of lawyers; assembling a bid on short notice can take days of intensive effort. Ropes & Gray, an early adopter of AI in the legal industry, aims with this tool to deliver expertise to more clients more quickly.
Capabilities of the new tool
The system, initially focused on material commercial contracts, performs several tasks:
- it scans thousands of pages to identify relevant documents;
- it automatically groups documents by themes and issues so relationships are reviewed rather than isolated files;
- it produces research summaries and a completed report in Ropes & Gray’s standard format;
- it inventories and categorizes the contents of a data room.
Lawyers can inspect the groupings and source documents to confirm completeness and accuracy before any client delivery.
Development and quality checks
In working sessions, Ropes & Gray lawyers compared the tool’s draft outputs against human-authored diligence reports from completed deals and adjusted the rubric and report format based on where the drafts matched or diverged from lawyer work. The firm designed the tool with a consistent query and framework that mirrors the team’s exact steps from initial review to higher-level analysis. Ropes & Gray characterizes the system’s output as comparable to the work of a capable junior-to-mid-level associate: a high-quality first pass on which lawyers can build.
Lawyer oversight and responsibility
Howard Glazer, Partner and Head of Practice AI at Ropes & Gray, emphasizes that lawyers must still review everything: “Our clients don’t pay us to review contracts. They pay us to make judgments based on what’s in the contracts.” Lichtenstein notes the importance of being able to stand behind 100% of the work delivered to clients, which is why the tool allows easy review of interim summaries and original documents. With the initial review automated, lawyers can spend more time refining analyses and crafting final judgments.
The firm has publicly stated an ambition for the tool to produce a detailed, issue-level diligence report in hours rather than the weeks such work has historically required. Ropes & Gray also cites the use of GPT‑6 Astra alongside their proprietary tool as enabling faster client problem-solving.
Lessons learned and future vision
Lichtenstein and Glazer share two practical lessons from building and using AI in legal workflows:
- Experiment broadly with AI: Lichtenstein describes using AI in many areas of daily life and professional work, learning both strengths and failure modes by regular use and iterating based on observed mistakes.
- Keep clients central: while the technology offers opportunity, the firm focuses on using AI only in ways that benefit clients, ensuring the tools are deployed to produce client value.
Looking ahead, Glazer expects the next stage of AI to help create collective knowledge within the firm. With 340 partners and 1,500 lawyers, Ropes & Gray sees value in making individual knowledge — from client meetings to past projects — more accessible across the firm. “That’s not just saving time. It’s not just saving money,” Glazer says. “It’s using the AI tools to directly create value, and that’s just amazing.”
Conclusion
Ropes & Gray’s collaboration with OpenAI aims to automate the initial, data-intensive pass of transactional diligence, yielding faster turnaround and consistent first-pass quality while preserving lawyer oversight and final responsibility. The firm hopes this will both shorten timelines and enable better use of firm-wide knowledge to benefit clients.
