AI Contract Review Takes Center Stage
AI contract review is making significant strides in the legal industry, as highlighted by recent developments involving leading platforms such as Claude and LegalOn. The conversation around legal AI was reignited this week after Nav Toor shared 12 detailed legal prompts for Claude, incorporating the names of top law firms like Latham & Watkins, Kirkland & Ellis, and Wachtell Lipton. This move followed Anthropic’s release of a legal plug-in, essentially a sophisticated prompt, which stirred debate across the legal tech community.
The use of renowned law firm names in AI prompts quickly sparked discussion among lawyers online. Many questioned whether referencing prestigious firms could influence the quality or style of the contracts generated by Claude. Some legal professionals expressed concerns about the implications for client work, emphasizing that while AI contract review tools are evolving, they still require careful oversight from human lawyers.
Testing Claude and GPT-5.4 for Contract Drafting
To assess the impact of these prompts, an experiment was conducted by running one through OpenAI’s latest GPT-5.4 model. The result was a comprehensive contract that appeared to satisfy the prompt’s requirements. However, as with all legal documents, the devil is in the details. The core question remains: can AI contract review generate contracts that are truly reliable, or do they risk missing critical nuances?
While using a prompt alone is not a substitute for a seasoned legal professional’s input, these experiments underscore the progress of AI in contract generation. For those interested, the full set of 12 prompts is available for further exploration—though experts advise caution before deploying them in actual client work.
LegalOn’s Analysis: GPT-5.4 Outperforms Previous Versions
LegalOn recently published an analysis comparing the performance of GPT-5.4 versus the earlier 5.2 release for contract review tasks. Their findings reveal notable improvements:
- Overall accuracy increased to 79.4% (up from 73.9%), a 5.5 percentage point improvement.
- GPT-5.4 reduced total errors by 21%, from 129 down to 102.
- Every contract type analyzed showed improvement, with NDAs seeing a 10 percentage point gain and MSAs an 8-point gain.
- Both precision and recall rose, resulting in fewer false positives and missed violations.
These results highlight a steady upward trajectory for AI contract review tools. As legal large language models (LLMs) continue to advance, their growing accuracy means less oversight is required, shifting the role of lawyers from hands-on drafting to high-level review and approval. If this pace continues, legal AI could soon reach accuracy levels comparable to experienced human practitioners.
Industry Adoption of Legal AI Platforms
The momentum in AI contract review is echoed by broader industry moves. The Financial Times has chosen Wordsmith as its enterprise legal AI platform, deploying it across legal, compliance, and company secretarial functions to drive efficiency and speed up access to legal support. Unlike point solutions, Wordsmith will manage everything from contract review and drafting to compliance and broader legal support.
Meanwhile, Sandstone has welcomed Kevin Keller, a seasoned legal and technology executive, to its board of advisors, underscoring the growing intersection of legal expertise and AI innovation.
Harvey, another legal AI powerhouse, has secured partnerships with the Dallas Mavericks, American Airlines Center, and Fulham FC, becoming the Official Legal AI Partner for the UK club. Additionally, Harvey and The LegalTech Fund have announced a joint initiative to invest in legal tech startups, aiming to foster greater accessibility and efficiency in the sector.
New Developments in AI-Driven Legal Tools
International law firm Herbert Smith Freehills Kramer is rolling out Legora, a general-purpose AI platform, across its operations. Centari, a deal intelligence platform, has expanded its capabilities with new products designed to simulate the reasoning of transactional lawyers and help clients extract strategic value from their data. AltaClaro and Verbit’s DepoSim, an AI-driven deposition simulator, has been selected by US law firm Taft for lawyer training purposes.
Upcoming Legal Innovators Conferences
The Legal Innovators conference series continues to be a focal point for legal tech advancement. Legal Innovators Europe will take place in Paris on June 24 and 25, while Legal Innovators California is set for June 10 and 11 in the Bay Area. These events offer a platform for law firms, in-house teams, and legal tech companies to connect, share insights, and drive the next wave of innovation in AI contract review and beyond.
Conclusion: The Future of AI Contract Review
The rapid evolution of AI contract review is reshaping legal workflows, with tools like Claude, LegalOn, and others setting new standards for contract accuracy and efficiency. As large language models continue to improve, their integration into legal practice will only deepen, offering both opportunities and challenges for legal professionals. Staying informed and engaged with these developments is essential for anyone navigating the future of legal technology.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.