When using AI in your internal investigation, make sure you protect the attorney-client privilege.

A lot of investigators are feeding interview notes, documents, even witness statements into AI tools to help organize an investigation.

That’s convenient, for sure, but ask yourself: where does that data go? Is that platform actually covered by your privilege log? Did outside counsel direct that use, or did an investigator just start using a public AI tool because it was faster?

You wouldn’t just hand out your interview notes to third parties who are not under privilege. That’s what AI can do, unbeknownst to you.

If the answer is unclear, you may have just created a discoverable record that sits outside the protection of attorney-client privilege or work product doctrine.

Opposing counsel or a regulator can ask what tools you used, what prompts you entered, and what the AI generated.

If that trail wasn’t properly walled off under privileged attorney-client use, you may have to produce it.

And there’s a second failure mode: shadow AI. Investigators using unapproved tools on their own, without telling anyone, to save time.

You won’t find out until it surfaces in a document request.

My advice: before your next investigation starts, define exactly which AI tools are approved, who directs their use, and how that use is documented and protected.

Get ahead of it, because opposing counsel is already asking the question.

The Ethics and Compliance Q and A show is produced by One Stone Creative.