Article details

ChatGPT - Benefits for Law Firms - Part 3 of 4 - Legal Contract Review

February 2023

ChatGPT is a state-of-the-art language model developed by OpenAI. It uses deep learning algorithms to generate human-like responses to text-based inputs. The model was trained on a large corpus of text from the internet, including websites, books, and other written materials, and has the ability to generate responses in a wide range of styles and formats, including natural language text, code, and other forms of structured data.

Related: ChatGPT - Benefits for Law Firms - Part 1 of 4 - Legal Research , ChatGPT - Benefits for Law Firms - Part 2 of 4 - Legal Document Drafting and Review , ChatGPT - Benefits for Law Firms - Part 3 of 4 - Legal Contract Review

ChatGPT has a number of applications, including natural language processing, conversation systems, language translation, text summarization, and document generation. It can be used to automate tasks such as customer service, content creation, and research.

Law Firms are using ChatGPT in a number of ways. In this 4 part series we will target different ways that Inherent, Inc. is helping Law Firms benefit from ChatGPT. In this articles we will focus on Legal Contract Review.

How does ChatGPT assist with Legal Contract Review?

ChatGPT was trained on a large corpus of text from the internet, which includes a wide range of information, including legal websites, legal databases, and legal textbooks. During its training, the model learned patterns and relationships within the text data, and it uses that information to generate responses to new text inputs.

Fine-tuning and Customizing ChatGPT: ChatGPT is trained using a variant of the Transformer architecture and a deep learning technique called unsupervised learning. While ChatGPT is powerful in its out of the box state, it's most effective, when it is customized and trained specifically for the organization that is using it. That is where Inherent, Inc. can come in to help build your custom application. Contact us here for more information.

How can you Train/Fine Tune ChatGPT to Review Contracts?

  • Provide examples: Provide ChatGPT with examples of legal contracts. These examples will help ChatGPT learn the language and style used. This data could include various types of legal contracts, such as employment contracts, lease agreements, and purchase contracts.

  • Identify key terms: Identify the key terms and concepts relevant in the contracts. This will help ChatGPT understand the specific terminology used. The training process would involve fine-tuning an existing language model on this legal text data, which would involve modifying the model's weights and biases to improve its ability to understand legal language and identify key clauses and concepts within contracts. The fine-tuning process would require carefully selecting appropriate hyperparameters and tuning the training process to optimize performance.

  • Create a dataset: Collect a dataset of documents and examples of the contracts, and label them with appropriate tags and categories.

  • Fine-tune: Use the dataset to fine-tune ChatGPTs language model to better understand the language and structure of the contracts. You may also need to develop additional post-processing steps or heuristics to improve the model's performance and ensure that it correctly identifies important legal concepts and clauses within contracts. This process can be complex and time-consuming, but it has the potential to greatly improve the efficiency and accuracy of contract review.

Here are examples of questions you can ask ChatGPT when reviewing a Contract

  • What are the key terms of the contract?
  • What are the rights and obligations of each party under the contract?
  • Are there any potential issues or risks with the contract?
  • How can the contract be terminated or renewed?
  • What happens in the event of a breach of the contract?
  • What are the governing law and jurisdiction of the contract?
  • What are the payment terms and methods under the contract?
  • Are there any warranties or representations made in the contract?
  • Are there any limitations of liability or indemnification provisions in the contract?
  • What are the confidentiality and intellectual property provisions in the contract

Here are a few examples of questions I asked ChatGPT about an RFP we received for Web Development

Q. What are the goals of the above RFP?

A. The goals of the above RFP (Request for Proposal) are to redesign Company Name's external website to improve its access, navigation, and information retrieval, while also centralizing content and optimizing workflow. Additionally, the goal is to create a robust web content management system solution that would allow for different tiers of administration, maintain or improve employee directory search functions, create a mailing list system, implement an improved site search function, integrate with current external applications, and add emerging technologies to encourage interaction/exchange with users. The website should also have an updated, fresh new corporate look and feel. The chosen vendor must also demonstrate the ability to provide site usability testing and debugging as well as staff training on the Content Management System.


Q. Is there a budget mentioned in the above RFP?

A. There is no information provided in the above RFP about a budget.


Q. What are the requirements of the RFP?

A. Improve site usage, navigation, and information retrieval - A robust, enterprise-ready web content management system solution - Maintain or improve employee directory search functions - Create a mailing list system for use with seminars, events, announcements that has cross-disciplinary capabilities and can be easily administered - Implement an improved site search function (develop or recommend) - Integrate with current external applications (Deltek Vision, eProject, WebEx, RecruitMax) - Event management and online calendar functionalities - Add emerging technologies to encourage interaction/exchange with users (blog commenting, mailing list expansion, event registration, client surveys) - Develop an updated, fresh new corporate look and feel

The RFP also specifies that the chosen vendor must demonstrate understanding and ability to provide services in site usability testing and debugging, staff training on the content management system, and ongoing maintenance and support.

The training process for ChatGPT can take several weeks or even months, depending on the size and complexity of the model and the amount of data used for training. Additionally, training large language models like ChatGPT requires significant computational resources, including GPUs and high-performance computing clusters.

If you are interested in learning about how Inherent & ChatGPT can help your Law Firm, please Contact Us for more information.