Worth Reading – The Hidden Dangers of AI Chat History

Rod echoes something I’ve been trying to talk to people about in the Copilot space for about a year now:v

Every conversation you have with an AI leaves a trail. Those chats feel private—like talking to a smart, non-judgmental friend who never gets tired. But the history of those conversations is often far less private than it appears. What you type, the questions you ask, the details you share, and even the patterns of how you think can be stored, analyzed, retained, and potentially exposed in ways most people never consider.

https://rodtrent.substack.com/p/the-hidden-dangers-of-ai-chat-history

There’s the obvious privacy issue with public AI tools, but people often overlook another reality: at work, someone likely has the same access to your AI interactions as they do to your email and Teams messages. As a company using AI, you’re also responsible for all that data. You must protect it, keep it private, and, if required, collect it in response to a lawsuit or investigation.

Are you prepared to do that?

I’ve done webinars and newsletters on this topic with Microsoft Copilot. If you’re interested in learning more or hiring me to train your team on discovering Copilot interactions, I’ve done that for a few folks on the side. Let’s talk. 

Follow these topics: LitigationSupport