Search This Blog

Powered by Blogger.

Blog Archive

Labels

Showing posts with label AI Banned. Show all posts

Locking Down ChatGPT: A User's Guide to Strengthening Account Security

 



OpenAI officials said that the user who reported his ChatGPT history was a victim of a compromised ChatGPT account, which resulted in the unauthorized logins. OpenAI has confirmed that the unauthorized logins originate from Sri Lanka, according to an OpenAI representative. According to the user, he logged into his ChatGPT account from Brooklyn. 

In the leaked private conversation, the employee appeared to be troubleshooting an app; the name of the app and the location where the problem occurred were also listed. According to reports in ArsTechnica, there is a report that private conversations on ChatGPT were leaked. 

Among the details leaked are login credentials and other personal information of unrelated users. The report also provided screenshots submitted by the alleged hacker of the account. Several screenshots have been shared, including several pairs of passwords and usernames that appeared to be related to a support system that is used by pharmacy employees to assist with prescription drug ordering. 

Any personal data you share in your chat history can be accessed by hackers if your OpenAI account is hacked. Even though this makes perfect sense, it is very strange that you can gain access to information from other compromised accounts, especially in the context of security threats. 

When using OpenAI, you need to make sure you use a strong password to secure your ChatGPT history as it does not provide multi-factor authentication. To ensure that your OpenAI account is secure, you will need to follow basic security measures similar to those that you would take with any other online account. 

Almost everybody does not want to memorize a long passphrase, which includes letters, numbers, symbols, and cases, not to mention a different passphrase for every account. This is why there are password managers out there. 

It is important to note that if you do not use the built-in password manager on your phone, laptop or browser, you will want to visit the Best Password Managers page for help in choosing the best password manager for your situation. If you suspect any account may have been compromised, you should change your password immediately to a long, unique passphrase. 

The user, Chase Whiteside, has since changed his password but is not convinced that his account has been compromised. According to him, he used a password with nine characters, including upper-case letters and lower-case letters, plus special characters, as well as special characters, but he said he didn't use it anywhere else but for his Microsoft account. 

When he briefly stopped using his account on Monday morning, the chat histories of other people appeared all at once. Hence, OpenAI's explanation suggests the initial suspicion that ChatGPT leaks chat histories to unrelated users may not be accurate. 

Despite these shortcomings, the report notes that the website does not contain an option for users such as Whiteside to protect their accounts using two-factor authentication or track details such as the IP address of their current and recent logins - both of which have been standard across most major platforms for some time. 

According to a November paper published in the journal Science, researchers showed how queries were used to prompt ChatGPT into divulging information that was contained within the material that was used to train the ChatGPT large language model, such as email addresses, phone numbers, fax numbers, and physical addresses. 

Several companies, including Apple, have restricted the use of ChatGPT and similar services by their employees for fear of sophisticated or proprietary data leaks among employees. There are a number of reasons for this restriction.