AI Forensics vs. Traditional Digital Forensics: Examining AI Systems in Litigation

This session explores the rapidly expanding landscape of AI-related litigation, examining how the average large company now uses over 35 AI systems and algorithm-related cases have increased 300 percent. Participants will learn to distinguish between predictive systems, computer vision, language processing, generative AI, and autonomous systems, understanding how each type presents unique legal challenges.

This session examines how AI forensics differs fundamentally from traditional digital forensics, focusing on system behavior and decision processes rather than static file recovery. Attendees will learn the four-phase AI forensics framework covering preservation, system identification, testing and validation, and documentation while understanding why AI’s probabilistic nature requires specialized analytical approaches.

A brief intermission allowing participants to refresh before continuing with the remaining sessions. Use this time to reflect on concepts covered and prepare questions for upcoming discussions.

This session addresses the critical sources of bias in AI systems including data bias, developer bias, and application bias that can affect fundamental rights in litigation. Participants will explore the ‘black box problem’ and learn techniques for examining AI-derived evidence when deterministic answers are not possible.

This session provides practical strategies for presenting AI-related evidence to fact finders in both civil and criminal proceedings. Attendees will learn discovery planning techniques, how to address trade secret objections, and the importance of building multidisciplinary expert teams early in case development.