AI tools like ChatGPT, Claude, and others can be incredibly useful, and we expect they are only going to become more common in day-to-day business. But there is one area where a lot of caution is warranted right now: putting sensitive information into consumer AI tools.
A recent decision out of the Southern District of New York, United States v. Hepner, is a good example of why. In that case, the court held that a litigant’s written exchanges with Claude were not protected by attorney-client privilege or the work-product doctrine. The court’s reasoning was pretty straightforward: Claude was not the person’s lawyer, the information had been shared with a third-party platform, and the prompts and outputs were not created at the direction of counsel.
Simply put, information entered into a public AI tool—and the responses it generates—may be treated a lot like emails, texts, or other third-party records that can be obtained in discovery.
That does not mean businesses should avoid AI altogether. Far from it. These tools can be extremely helpful. But this is still a developing area, and until the law becomes more clear, it makes sense to be careful—especially when legal strategy, confidential business information, or other sensitive matters are involved.
For now, clients should assume that prompts, chat histories, uploaded documents, and AIgenerated outputs may be discoverable. That can include things like:
- Legal questions
- Strategy brainstorming
- Draft arguments
- Internal investigations
- Compliance issues
- Other confidential business information
The key takeaway for you? If you would not want an opposing party to see it, do not put it into a consumer AI tool!
For example, say you are considering breaching a contract. You type into ChatGPT, “What happens if I breach a contract?”
Later, litigation occurs and this message is discovered. That seemingly simple question you asked an AI certainly suggests that you believed you were breaching the contract and undermines any arguments to the contrary.
A few practical precautions can go a long way: review the terms and privacy settings of any AI platform before using it, avoid entering confidential or legally sensitive information into public-facing tools, and talk with counsel before using AI in connection with disputes, investigations, compliance matters, or legal strategy.
Let us handle it! If you have questions about how to use AI tools more safely in your business, we would be glad to help.
For further information, or assistance with planning for the impact of these rules on your business and employees, please contact Gardner Skelton.
Client Alert Authors:
Ben Sarver; Marketing Director
Kyle McClain; Senior Paralegal
