In our second installment on the legal implications of generative AI, we will talk about electronic discovery issues associated with AI. Even prior to the commencement of legal proceedings, it is vital for an organization to understand whether and how AI tools are used within the organization. Companies should be aware of who might be using AI models as well as where datasets fed into the AI model, prompts used to engage with the model, AI algorithms, and model outputs are stored, and for how long. Understanding the framework surrounding a company’s AI model can help counsel develop a well-considered discovery plan once litigation is anticipated.

Type of AI Model and Its General Impact on Discovery

To begin, it is critical to assess the type of AI model your organization is using and who is using it in order to thoughtfully evaluate the discovery implications surrounding the use of the model and its output. Broadly, AI can be categorized as either an open source or a closed loop system, each presenting distinct, though sometimes overlapping, legal considerations especially with respect to preservation and discovery issues.