Horrified users: Conversations with ChatGPT may have become visible to others

A bug in ChatGPT has caused private user requests to be briefly visible in Google Search Console.

This is an analytics tool that website owners use to track traffic on their websites.

The bug meant that certain queries sent to the chatbot were published in a way that could compromise user privacy.

OpenAI has stated that the bug was fixed as soon as it was identified.

According to the company, only a small number of queries were affected.

OpenAI explained that the issue was due to a technical error where certain user queries were inadvertently sent to Google Search Console.

Here they could potentially be seen by others with access to the relevant analytics panel.

The company has not disclosed what types of queries were affected or how long they were visible in Google Search Console.

Expert points to possible pattern

Jason Packer, owner of US analytics firm Quantable, has suggested that the incident may reflect a practice on OpenAI’s part where the company tries to pull information from Google to improve chatbot responses.

However, it has not been confirmed that this practice directly caused the current ChatGPT bug.

The error once again highlights the challenges associated with the use of AI chatbots in relation to data security.

ChatGPT and similar technologies have been the subject of criticism in the past, including legal requests for access to users’ conversation histories and privacy issues.

OpenAI has maintained that the bug has been fixed and that only a limited number of users were affected by the leak.

Nevertheless, the incident has brought renewed attention to whether full privacy can be expected when using AI tools like ChatGPT.

Several experts warn that sharing sensitive or personal information with this type of technology should be avoided precisely because accidental data leaks cannot be completely ruled out.

Efforts are underway to improve data security standards at OpenAI and ChatGPT.