On February 8th, it was reported that a hacker posted on the dark web forum, claiming to have the login credentials of 20 million OpenAI user accounts, and began to sell them publicly. OpenAI said that it has urgently launched an investigation, and preliminary investigation evidence shows that its system has not been attacked, but users still need to be alert to latent risks and take necessary security measures. An anonymous hacker posted Russian messages on the dark web forum, claiming to have access codes for more than 20 million OpenAI accounts, calling it a "gold mine", and also provided sample data containing email addresses and passwords. Security researchers verified the sample data and found that at least two of the email addresses were invalid, and other posts on the forum by the user were also related to stealing logs. The relevant posts have been deleted, casting doubt on the authenticity of the leak. OpenAI has issued a statement saying it is taking the matter seriously, but has not found any evidence that the OpenAI system was hacked. (Jin Ten)