Attackers abuse shared ChatGPT and Claude chats to spread malware

Attackers are abusing public ChatGPT and Claude shared chats to impersonate outage notices, install guides, and support pages.

Victims are often led in from search ads and trust the links because they appear to come from official domains.

Some chats use fake error pages rendered with ChatGPT and malicious Terminal commands in Claude to push infected downloads.

Security researchers say the tactic turns legitimate-looking AI conversation URLs into a new malware delivery channel.