Even the most casual observer of current news appreciates the projected use of artificial intelligence or “AI” to improve service delivery across a variety of industries. Artificial intelligence techniques such as natural language processing, machine learning, and deep learning have been commercially available for several years and are incorporated into some of the more widely adopted software applications (for example, Lex Machina which analyzes court dockets to predict case timing and judges’ penchants).
The more recent maelstrom involves the use of a species of AI called generative AI (GAI). GAI is a subset of deep learning (a sophisticated sub-field of machine learning) that, based on a user “prompt” (i.e., query or request), creates new images, computer code, video, audio, or natural language based on enormous amounts of data analyzed to train the algorithm. GAI is designed to generate content – it is not an editor, not an analytic tool (yet), and not a web browser. It doesn’t necessarily operate in “real time”: it may be basing its work off ingested materials that are years old. Failure to understand this limited functionality (which concededly GAI performs brilliantly) is the genesis of much of the criticism and concern about the use of these applications.
Here’s a list of a few of the questions that a client might ask about the use of AI and GAI. For these purposes, “AI” includes but is not limited to GAI.
Firm’s Use of AI
- Does the law firm utilize AI in its legal practice? If so, in which specific areas of law? Are there differences from practice area to practice area in how the AI is being used?
- Does the firm have a clearly defined policy for restricting the use of AI in specific cases or overall if necessary?
- How does the firm monitor compliance by its attorneys of AI and GAI use in accordance with the firm’s policy?
- What is the firm’s policy on use of personal accounts created in generative AI tools by firm attorneys?
- If the client already has an AI policy in place, will the firm adhere to that policy while providing services?
- What is the process for resolving disputes or challenges related to the use of GAI in the client’s legal matters? Does the law firm disclose to clients when GAI is used in their legal matters, and are clients given the option to opt out of its use if desired?
- Will AI tools be used in the specific project? How are AI tools integrated into the firm’s workflows?
- Can the client request the firm use a specific AI tool for a project or matter that the firm is working on?
- Are there any additional costs associated with the use of GAI, and how are these charges determined?
- Can the client request access to the GAI-generated content and information pertaining to their matters?
- What training and expertise do the law firm’s attorneys and staff have in working with GAI tools? What training of the lawyers/staff is in place to ensure safe use? Efficient use?
- If the firm is using GAI, what steps is it taking to mitigate risk related to the software supply chain, including both commercial and open source providers?
Benefits and Considerations
- What are the primary benefits and advantages of using AI in the firm’s legal services? Does the firm have any comparative data or metrics to support these conclusions?
- Are there any notable success stories or case studies where AI has significantly benefited clients?
- Has the law firm conducted any independent audits or evaluations of its AI systems to assess their effectiveness and legality?
- What safeguards are in place to ensure the accuracy and reliability of the AI tools?
- What steps are taken to keep the GAI technology and GAI-generated content up to date and in compliance with changing legal regulations?
- What steps are being taken to manage the long-term maintainability of AI tools, such as loss of access to the original training materials, loss of access to the specific model or tool, or changes and updates to the model or tool?
Risks and Compliance
- What are the limitations or potential risks associated with the use of AI in the legal work, and how does the firm address them?
- What measures are taken to protect the client’s sensitive information and data while using AI tools?
- How does the firm address potential biases that might be present in the AI models?
- What data is being used to train the GAI? What processes are in place to ensure the training data is being used in compliance with relevant property rights, laws, and regulations?
- How does the firm ensure compliance with legal and ethical standards while using GAI?
For several years, Thompson Hine has used AI-enabled tools in a variety of contexts. These include commercially available tools that rely on natural language processing and machine learning to analyze large quantities of documents and analytic tools that incorporate proprietary algorithms we have developed to analyze large quantities of data related to our delivery of legal services.
In keeping with our promise to the market of “Always Innovating,” Thompson Hine is assessing adoption of GAI tools to assist in our delivery of legal services to clients as well as in our internal business operations. We will strive to utilize large language model (LLM) and GAI tools as fully and as quickly as we can within necessary guardrails.
The authors developed this checklist based upon their experience and available resources regarding generative AI, including ChatGPT, which the authors used to generate ideas and suggested text.
This article may be reproduced, in whole or in part, with the prior permission of Thompson Hine LLP and acknowledgement of its source and copyright. It is intended to inform clients about legal matters of current interest. It is not intended as legal advice. Readers should not act upon the information contained in it without professional counsel.
This document may be considered attorney advertising in some jurisdictions.
© 2023 THOMPSON HINE LLP. ALL RIGHTS RESERVED.