Doing Business in Florida Blog
Could Your AI Chats Become Evidence? What Businesses Should Know
October 8, 2026
Artificial intelligence tools such as ChatGPT, Microsoft Copilot, Google Gemini, Claude, Harvey and Siri have quickly become part of everyday life. People use them to research unfamiliar subjects, explore health concerns, create presentations and develop plans of action.
But as with almost every human activity, using AI tools carries legal ramifications. An AI conversation may preserve a user’s questions, assumptions and intentions—along with any confidential information entered into the platform. If a dispute arises, that chat history could become relevant evidence.
Your AI Conversations May Not Be Privileged
Communications with your lawyer may be protected from third-party disclosure by the attorney-client privilege. Conversations with an AI chatbot about potential legal claims, defenses or strategies generally do not receive that protection merely because they concern a legal issue.
In many respects, asking a chatbot for legal guidance may be more like chatting, texting or emailing with a friend or neighbor. The fact that the conversation involves legal questions does not, by itself, make it privileged.
Privilege and work-product protections are fact-specific. Businesses and their employees should not assume that an AI conversation is protected or that using a chatbot is a substitute for communicating confidentially with counsel.
Are You Sharing Confidential Information?
Users frequently provide background information so an AI tool can generate a more relevant response. But that background may include:
- Confidential business information;
- Trade secrets or proprietary processes;
- Customer or employee data;
- Financial information;
- Contract terms;
- Details about an internal investigation; or
- Communications with legal counsel.
Do you know how the AI provider stores that information, who may be able to access it or whether it may be used to improve its models?
The answers vary depending on the platform, account type, contractual protections and privacy settings. An enterprise product operating under negotiated confidentiality and security terms may present different risks from a publicly available consumer tool.
Before permitting employees to use AI for sensitive work, businesses should understand the platform’s terms, settings and safeguards. Entering protected information into an unapproved tool could expose it to unintended access or complicate efforts to preserve its confidentiality.
AI Chats Could Become Evidence
You may express legal concerns through AI while hoping to identify a path to a successful outcome. But your prompts and your actions in response to the answers could become relevant in a later proceeding.
Consider a business owner who asks an AI chatbot to develop a plan for removing a business partner on questionable grounds. The chatbot warns that the grounds may not be legally sufficient but nevertheless provides a roadmap. The owner then follows it.
If a dispute arises, that conversation could potentially be used to show:
- What the business owner knew;
- Whether the owner recognized weaknesses in the proposed grounds;
- Which alternatives the owner considered;
- The owner’s intent or motive; and
- Whether subsequent actions followed the AI-generated plan.
Not every AI conversation will be discoverable or admissible. Those questions depend on factors such as relevance, proportionality, privilege, authenticity and whether the information is within a party’s possession, custody or control. Nevertheless, AI conversations may receive the same scrutiny as emails, text messages and workplace chats.
AI Chat Histories May Need to Be Preserved
Once litigation is pending or reasonably anticipated, a business may be required to take reasonable steps to preserve potentially relevant evidence. Depending on the dispute, that obligation could include:
- Prompts and chatbot responses;
- Documents uploaded to the platform;
- Saved or exported conversations;
- Account histories; and
- Related records maintained through connected applications.
Deleting a conversation may not eliminate it. Copies or related records could remain on personal devices, company systems, connected applications or the provider’s infrastructure.
AI platforms should therefore be considered when issuing litigation holds and identifying potential sources of electronic evidence. If relevant information is lost after a preservation duty arises, the business may face remedial measures or, in appropriate circumstances, sanctions.
Practical Steps for Businesses
AI tools offer significant benefits, and businesses do not necessarily need to prohibit their use. They should, however, establish appropriate safeguards:
- Adopt an AI-use policy. Identify approved tools, acceptable uses and information employees may not enter.
- Protect sensitive information. Do not place confidential information, trade secrets, personal data or legal communications into an unapproved platform.
- Review provider terms and settings. Understand how prompts and uploaded materials are stored, accessed, used and deleted.
- Train employees. Make clear that AI conversations are not necessarily private or privileged and could become evidence.
- Address AI in litigation holds. When a dispute arises, determine whether relevant AI conversations and related records must be preserved.
- Consult counsel before using AI for sensitive legal questions. Legal strategy, anticipated disputes and significant employment or corporate decisions should be discussed confidentially with an attorney.
The Bottom Line
AI has seemingly universe-bending capabilities, but its use carries meaningful risks. Before entering a prompt, consider whether you would be comfortable seeing both your question and the chatbot’s response displayed in a deposition, regulatory investigation or courtroom.
Like an unfortunate email or text message, an AI conversation could come back to bite you. Businesses should remain mindful of what they share, what they ask and how they act on the information an AI tool provides.
Berger Singerman’s Dispute Resolution Team is available to help businesses assess the legal risks associated with AI communications, develop appropriate policies and preservation procedures, and respond when AI-generated records become relevant to a dispute. For more information, please contact Anthony Carriuolo or your Berger Singerman attorney.
This article is provided for informational purposes only and does not constitute legal advice. The information should not be relied upon as a substitute for advice from qualified legal counsel. No attorney-client relationship is created by this article. Readers should engage and consult counsel regarding their specific circumstances.
