Sullivan & Cromwell
A leading United States law firm founded in 1879 and headquartered in New York. It is known for corporate mergers and acquisitions, litigation, and regulatory practice, and in recent years has actively explored the application of generative AI in legal services.
Core Facts
Timeline (last 90 days)
Sullivan & Cromwell's practice demonstrates the potential of combining general-purpose programming tools with domain knowledge, and law firms can use tools like Codex to encapsulate their own precedent libraries, playbooks and business judgment into custom applications.
Sullivan & Cromwell lawyers input the draft contract, relevant precedents, the playbook representing the client's position, and market conventions into the AI model together, and the model outputs a memo listing identified issues plus revision markups.
When a law firm uses Codex, lawyers familiar with the business process need to describe the requirements, the model generates a script or application skeleton, and then technical staff or the lawyers themselves debug it.
According to the video description, Sullivan & Cromwell's use of Codex for contract review is far faster than lawyers doing it alone.
A video released by OpenAI showed Sullivan & Cromwell using Codex to transform contract review from lawyers reading word by word into a new process of AI-generated memos plus revision markups.
The practice of law firms collaborating directly with model vendors to build tools tailored to their internal workflows is not common in the legal industry.
The specific data isolation scheme between Sullivan & Cromwell and OpenAI has not yet been made public.
Sullivan & Cromwell partnered with OpenAI to launch an AI tool called Agreement Analyzer, used to help lawyers quickly identify potential issues in draft contracts.
All Facts (9)
Deeper analysis by AI code review tools means handling larger context windows, multi-round reasoning, and potentially invoking models with larger parameter counts, which significantly increases token consumption and response time.
75%UnverifiedSullivan & Cromwell's practice demonstrates the potential of combining general-purpose programming tools with domain knowledge, and law firms can use tools like Codex to encapsulate their own precedent libraries, playbooks and business judgment into custom applications.
50%UnverifiedSullivan & Cromwell lawyers input the draft contract, relevant precedents, the playbook representing the client's position, and market conventions into the AI model together, and the model outputs a memo listing identified issues plus revision markups.
50%UnverifiedWhen a law firm uses Codex, lawyers familiar with the business process need to describe the requirements, the model generates a script or application skeleton, and then technical staff or the lawyers themselves debug it.
50%UnverifiedAccording to the video description, Sullivan & Cromwell's use of Codex for contract review is far faster than lawyers doing it alone.
50%UnverifiedThe practice of law firms collaborating directly with model vendors to build tools tailored to their internal workflows is not common in the legal industry.
50%UnverifiedThe specific data isolation scheme between Sullivan & Cromwell and OpenAI has not yet been made public.
50%UnverifiedA video released by OpenAI showed Sullivan & Cromwell using Codex to transform contract review from lawyers reading word by word into a new process of AI-generated memos plus revision markups.
50%UnverifiedSullivan & Cromwell partnered with OpenAI to launch an AI tool called Agreement Analyzer, used to help lawyers quickly identify potential issues in draft contracts.
50%