110,000 private chats went public. The worst one? A plot against the Amazon.
A shocking leaked ChatGPT conversation exposed a lawyer for a multinational energy firm seeking advice on how to forcibly remove an Amazonian indigenous community to build a hydroelectric dam.
The attorney asked how to secure the “lowest possible price” for the land, claiming the tribe didn’t understand its market value.
The chat was among 110,000 private conversations accidentally made public due to an OpenAI sharing feature flaw.
Other exposed chats included an Egyptian criticising their president (potentially putting them at risk), a domestic violence survivor’s escape plan, and inappropriate content involving minors.
The breach mirrors Meta’s similar April blunder when its AI chatbot publicly displayed users’ private conversations.
We just removed a feature from @ChatGPTapp that allowed users to make their conversations discoverable by search engines, such as Google. This was a short-lived experiment to help people discover useful conversations. This feature required users to opt-in, first by picking a chat… pic.twitter.com/mGI3lF05Ua
— DANΞ (@cryps1s) July 31, 2025
While the lawyer’s request may have been testing ChatGPT’s limits, it highlights AI’s potential for enabling exploitation.
The incident raises urgent questions about tech companies’ responsibility when their tools facilitate harmful behaviour, revealing how even advanced AI can become dangerous without proper safeguards.
As AI becomes more powerful, this leak underscores the need for stronger ethical protections.
