
Stop talking to ChatGPT as if you’re talking to your friend —because it’s not.
The introduction of AI has led many people to see these tools as their personal therapists, lawyers, and friends.
Talking to Meta or other AI tools like a friend has been gaining traction, especially on social media.
Many AI users believe that whatever they discuss with Meta or other AI tools is private and protected. But Sam Altman, CEO of OpenAI, says that’s not the case.
Altman confirmed that ChatGPT inputs are not protected by legal privilege.
So, if you’re thinking about telling ChatGPT or any AI tool your darkest secret—maybe you’ve committed murder or you’re planning one—don’t. Whatever you tell AI can potentially be used as evidence in court.
Speaking to Thando on Drive with Thando, Megan Harrington-Johnson, managing partner at HJW Attorneys, said OpenAI can be compelled to disclose conversations between you and the platform.
This is because OpenAI is not governed by the same regulations as law firms, where you are protected by legal professional privilege.
“The issue is that there’s no regulation at all,” said Harrington-Johnson.
Recently, ChatGPT users discovered that their conversations were being indexed by Google.
These chats ranged from emotional breakdowns to business struggles—all unintentionally shared.
Wondering how this happened? ChatGPT introduced a “share” button, which made shared conversations discoverable. Fortunately, this feature can be turned off.
After users complained about their private chats appearing in Google search results, OpenAI removed the “discoverable” setting—calling it a “short-lived experience.”