ChatGPT and Other AI Tools Used to Host Fake Outage Pages

This story caught my attention because, at first glance, it doesn’t make a lot of sense. When most people think of a ChatGPT share link, they think of a snapshot of a conversation. You share the conversation, someone opens the link, and they see what was discussed at that point in time.

According to recent reporting, threat actors have been abusing AI sharing features to distribute fake outage notices and malware. In the reported cases, victims were directed to legitimate ChatGPT share links containing content that appeared to be an official service notice. Because the page was hosted on a real ChatGPT domain, some users assumed the message was legitimate and followed instructions to download software.

The important point is that OpenAI was not hacked, and the attackers did not create a real ChatGPT outage page. Instead, they created content that looked like an official service notice and published it through the sharing feature. Victims saw a legitimate ChatGPT URL and trusted what they were seeing.

This attack highlights an important security lesson. We often tell people to check the website address before trusting a page. That remains good advice, but being on a legitimate domain does not automatically mean that every piece of content hosted there is trustworthy. Many platforms allow users to create and publish their own content, and that content may be misleading or malicious.

If you encounter one of these pages, you can report it to the platform involved. However, don’t wait for a response before protecting yourself. Close the page, avoid downloading anything, and verify any claimed outage through official sources.

A useful comparison is YouTube. Anyone can upload a video to YouTube, and the video is hosted on youtube.com. That does not mean YouTube endorses the video’s contents. The same idea applies to shared AI conversations. A shared conversation may be hosted on a legitimate AI platform, but the information within it may not be official.

Another warning sign is urgency. Attackers often try to convince people that they need to act immediately to restore service, verify an account, or install an update. Legitimate outages generally do not require users to download software from a shared conversation link. Outages will keep you on the domain with a page that explains what is happening and let you know if there is anything you need to know.

The safest approach is simple. If ChatGPT, Claude, Gemini, or another AI service appears to be experiencing problems, do not download software from a message claiming to fix the issue. Instead, close the page and go directly to the provider’s official website or status page to verify the information.

AI tools continue to grow and evolve with updates and improvements. As they do improve and grow, attackers will continue looking for ways to abuse the trust people place in well-known platforms. This incident serves as a reminder that we should evaluate the content we see, not just the domain name where it is hosted.

While this incident involved ChatGPT share links, the same principle applies to forums, social media platforms, file-sharing sites, video platforms, and other services that allow users to publish content. A trusted platform does not automatically make every piece of content on that platform trustworthy.

Read the full article by reading ChatGPT share links abused to host fake outage pages to deliver malware from Bleeping Computer.