Generative AI chatbots have quietly become part of how people process workplace conflict. A complainant may talk through an incident with an AI tool before deciding to report it. A respondent may use one to draft a response or think through their side of the story. A witness may describe what they saw the same way they’d text a friend. Workplace investigators who don’t think to ask about these conversations may be missing a source of evidence—one that is often contemporaneous, candid, and increasingly discoverable in litigation.
Courts Are Treating AI Chats Like Any Other Business Record
A growing body of case law confirms that interactions with AI chatbots are electronically stored information (ESI), subject to the same discovery rules as email, text messages, and Slack. In Fortis Advisors LLC v. Krafton, Inc., the Delaware Court of Chancery relied on the CEO’s ChatGPT interactions as it would any internal communication.
Elon Musk’s lawsuit against OpenAI offers a preview of what this looks like in practice. Personal texts, emails, and a digital diary belonging to OpenAI executives were pulled into the litigation, prompting commentators to compare AI chat logs to a diary, although diary records are what someone chose to write down, while a chatbot is built to draw more out of the user. The same dynamic can apply in an investigation: a chatbot conversation with a complainant, respondent, or witness may capture a more detailed, evolving account than anything the person put in writing elsewhere.
Criminal cases are starting to follow a similar pattern. In U.S. v. Heppner, 2026 WL 436479 (S.D.N.Y. Feb. 17, 2026) , the United States District Court for the Southern District of New York ruled that a defendant’s conversations with an AI chatbot were admissible, even though the individual argued that he used the tool to prepare for a conversation with counsel. Chatbot logs have also reportedly surfaced as evidence in criminal matters involving arson and homicide investigations.
The Role of the Workplace Investigator
Workplace investigations turn on credibility and a complete factual record—and that record increasingly includes AI conversations that complainants, respondents, and witnesses may not think to mention. A complainant may have used a chatbot to process what happened before filing a report. A respondent may have asked an AI tool to help draft a rebuttal or talk through a defense. A witness may have discussed the events with a chatbot the same way they might text a friend. These conversations can contain timestamped, contemporaneous accounts of what a party believed and said—often before positions harden into formal statements—making them potentially significant evidence that a thorough investigation should not overlook.
Because these chats are generally not protected by any special privilege, and because platforms may retain them even after a user believes they’ve been deleted, they represent both an evidentiary opportunity and a preservation risk. If litigation is reasonably anticipated, the duty to preserve extends to relevant AI chat logs held by any party to the investigation, not just company records.
Practical Steps
Ask about the employer’s AI use policy and practice as a standard part of intake. When interviewing complainants, respondents, and witnesses, investigators should routinely ask whether they discussed the matter with an AI chatbot and treat an affirmative answer as they would a reference to a text thread or journal entry.
Include AI conversations in the scope of evidence requests. Document preservation and collection requests directed at the parties should explicitly reference chatbot histories and exported conversation logs, not just emails and text messages.
Employers can consider updating litigation hold procedures. Once litigation is reasonably anticipated, holds issued to relevant individuals should specifically instruct them to preserve AI chat logs, including any conversations they may have already deleted, given that many providers retain data beyond user-initiated deletion.
Assess authenticity and context carefully. As with any digital evidence, investigators should consider chain of custody, account ownership, and the circumstances surrounding an AI conversation before relying on it in findings.
Coordinate early with IT and outside counsel on how to request, collect, and authenticate AI conversation data from company systems or personal accounts relevant to the investigation.
The Bottom Line
Courts are not writing new rules for artificial intelligence—they are applying familiar discovery and preservation principles to a new medium. For workplace investigators and the HR executives and attorneys who rely on their work, the practical implication is straightforward: build AI-use governance into your investigation and retention protocols.