Are AI chats discoverable?

Are AI Chats Discoverable in Legal Proceedings?

As artificial intelligence becomes more embedded in legal workflows, a pressing question emerges: can your AI chat conversations be used against you in court? The short answer is yes—they may very well be discoverable, even if created in anticipation of litigation.

Recent legal developments suggest that prompts entered into AI tools and the responses generated aren’t automatically shielded from discovery. This includes scenarios where lawyers use AI to draft arguments, summarize case law, or brainstorm strategy. If these interactions occur through third-party platforms, the data may fall outside the protection of attorney-client privilege, especially if the tool’s privacy policy allows data collection or sharing.

Courts are increasingly looking at the terms and privacy policies of the AI services in question to determine whether confidentiality applies. For instance, if an attorney uses a widely available AI chatbot that stores inputs for training purposes, those exchanges could be considered accessible to others—and therefore not truly private.

The takeaway? Legal professionals must be cautious about what they type into AI systems. Treating AI chats as confidential should no longer be assumed. Firms are encouraged to adopt clear AI usage policies, opt for secure and transparent platforms, and consider contractual safeguards when possible.

Just as you wouldn’t discuss a privileged strategy with a colleague in a crowded elevator, the same discretion should extend to AI interactions. As the legal landscape adapts, so too must the habits of those within it. The convenience of AI comes with responsibility—and oversight.