Meta Thwarts ChatGPT Malware Campaigns and Scams

Meta has taken down over 1,000 malicious URLs that used OpenAI's ChatGPT to propagate about 10 malware families since March 2023. The malware campaign, dubbed Ducktail, used fake ChatGPT web browser extensions to steal users' Facebook account credentials and run unauthorized ads from hijacked business accounts. Meta has also uncovered another strain of malware, NodeStealer, capable of stealing cookies and passwords from web browsers to compromise Facebook, Gmail, and Outlook accounts. The attack chains primarily target personal accounts of users connected to business pages and advertising accounts on Facebook. Meta is launching a new support tool to help users identify and remove malware and enable businesses to verify connected Business Manager accounts.