← the record
AIAAIC-1120

ChatGPT leaks user conversations, personal information

ChatGPT allowed some users to see other users' conversations and personal information, suggesting OpenAI had access to user conversations, and calling into question the company's privacy duty of care. Users had shared images of chat histories that they said were not theirs o n social media sites Reddit and Twitter, prompting a flood of complaints. Initially, the incident appeared confined to user conversations, which OpenAI states in its ChatGPT FAQs are reviewed - most likely to improve its systems on an ongoing basis. However, the company later confirmed that some users' first and last names, email addresses, payment addresses, parts of credit card numbers and credit card expiration dates had also been exposed. OpenAI disabled the chatbot to fix the error, and said that users had not been able to access full conversations. System 🤖 ChatGPT Documents 📃 OpenAI (2023). March 20 ChatGPT outage: Here's what happened Operator: Developer: OpenAI Country: USA Sector: Technology Purpose: Generate text Technology: Chatbot; Generative AI; Machine learning Issue: Privacy

Date it happened
2023-03-01
Product, system or model
ChatGPT
Where this came from
Share this incident
XLinkedInFacebookWhatsAppEmail
Attribution

This incident was imported from AIAAIC and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.