ChatGPT Empty Room Prompt Exposes User Data

A viral trend asks ChatGPT to furnish an empty room using what it remembers about the user, revealing stored personal details.

Updated on Nov. 8, 2026 in Cybersecurity

Isometric editorial illustration of a bare gray room furnished with a chair, stacked boxes, a teacup, a scarf, and a potted fern under a pendant lamp.
ChatGPT's viral empty-room prompt asks the chatbot to furnish a room using what it remembers about users, exposing stored personal details shared publicly on X. AI Illustration. Upload story photo >

Live Poll

Do you trust AI chatbots to keep your personal information private?

A viral prompt asks ChatGPT to fill an empty gray room with everything it knows about a user, and the furnishings reflect past conversations. Images from the trend are being posted publicly on X.

Why it matters

The trend shows how saved memories and chat history can surface personal details such as names, preferences, and goals when shared publicly. Users can view, delete, or switch off ChatGPT memory in the Personalization settings.

The trend relies on saved memories and chat history reference, which Plus and Pro subscribers have enabled even when nothing was saved. Earlier this year, developers found crypto wallet data in Claude chats that Google had indexed through shared links.

The players

OpenAI

OpenAI is the developer of ChatGPT, which lets users view, delete, or switch off its memory feature in the Personalization settings.

Meta

Meta owns Facebook and was ordered by a New Mexico judge to pay a $25 billion penalty over its handling of user data.

The details

The prompt pairs a plain image of an empty gray room with a single instruction, and a popular crypto account on X urged followers to try it and post the results. Temporary Chat neither reads nor updates ChatGPT memory, offering users a way to limit exposure.

Timeline

  1. Earlier this year: Developers found crypto wallet data in Claude chats indexed by Google.

Big Picture

The ChatGPT memory exposure follows the pattern of personal data leaking from AI chat tools, as developers previously found crypto wallet data in Claude chats indexed by Google through shared links. Both cases show that data users assume is private can become visible once it is shared or stored by the platform.

Users who have enabled saved memories or chat history reference should review their Personalization settings to delete or disable stored details. Crypto holders face heightened phishing and targeted scam risk if personal details are exposed.

The takeaway

Reviewing and clearing ChatGPT memory and avoiding public sharing of personalized chat images can limit how much personal information others can see.

Further reading

Explore more on Cybersecurity to see how AI platforms handle stored personal data.

Source note: This article includes information reported by BeInCrypto.

Live Poll

Do you trust AI chatbots to keep your personal information private?