ChatGpt's artificial intelligence is not free from technical and privacy problems.
In the past few hours, the popular software has stopped working for everyone.
The disservice was caused by the decision of the developer, OpenAI, to temporarily block the platform due to a bug, a flaw, which exposed the titles of user conversations.
The content of the conversations, as OpenAI specified to Bloomberg, was not disclosed.
After intercepting the problem, the company, in which Microsoft has invested heavily in recent months, has blocked access to the chatbot to prevent the privacy error from spreading further.
Before being taken offline, on the main page of ChatGpt, instead of seeing the history of the titles of one's chats with the AI, one could read those, random, of other surfers.
The company did not provide details on the criticality, with a spokesperson who limited himself to declaring the problem concerning privacy, at least as regards the questions posed to the software.
To avoid further issues, even after the resolution, user history remained unavailable, resulting in the inability to access previously asked questions to the AI.
The ChatGpt status page specifies that OpenAI is still working to restore everything.
In recent days, the organization has unveiled Gpt-4, the new generation of the model at the base of ChatGpt, which can also analyze images uploaded by users, always providing answers in text form.