A recent federal ruling highlights a growing risk for employers that use AI tools to capture notes, summarize meetings, or analyze sensitive workplace issues. In United States v. Heppner, the court held that documents created using Anthropic’s Claude AI tool—and later shared with legal counsel—were not protected by the attorney-client privilege.client privilege. 

The court’s reasoning was straightforward: AI tools are not attorneys, do not owe professional duties, and expressly disclaim confidentiality in their terms of service. As a result, materials generated through such tools remain unprivileged, even if later provided to legal counsel.  

For employers, this has direct implications for AI notetaking tools used in HR, management, or compliance contexts. Notes generated by AI during discussions about terminations, workplace complaints, or legal exposure are likely discoverable in litigation. This means that employees, HR teams, or managers using notetaking tools like CoPilot, ChatGPT or Claude to memorialize or analyze employee-centric topics may be creating evidence that can later be used against the company. 

To help mitigate the risk associated with the use of AI, including notetaking tools, employers should consider the following best practices:  

1. Keep sensitive legal matters out of commercial AI tools. Ensure your AI policy documents make clear that employees should not use generative AI platforms to analyze potential legal risk, create/discuss HR strategy, or create documentation related to disputes, investigations, or employment claims. 
2. Assume AI-generated content (including meeting notes) is discoverable. When it comes to AI always operate on the assumption that AI prompts, outputs, and conversation histories are not confidential and may surface in litigation, just like any other unprivileged document.  
3. Include clear use restrictions in your AI Policy and related documents. The use of AI should be clearly restricted to ensure it is not used in ways that might compromise the privacy and security of company data or infringe upon individuals’ rights. Specifically, it should be clear that AI notetaking tools are not to be used during internal or external meetings addressing employee-relations issues (e.g., performance evaluations, disciplinary meetings, mental health and wellness check-ins), legal and compliance discussions, executive strategy sessions, sensitive negotiations and contract discussions.
4. Train HR and managers on AI privilege risks. Because employees are increasingly turning to AI tools for analysis and documentation, employers should ensure all employees understand the risks associated with AI use. Specifically, that AI is not a substitute for legal advice and does not create protected communications.