Azerbaijan’s National Cybersecurity Agency has warned about a phishing campaign that uses fake pages resembling ChatGPT and other artificial intelligence services to steal login credentials and one-time security codes.
The fraudulent portals impersonate services including ChatGPT, Gemini, Claude, Perplexity, Meta Muse and Manus. Victims may be offered tools to connect business accounts, manage advertising campaigns or analyze spending.
After a user clicks through, the page displays a fake login panel designed to look like a browser window. Information entered into the panel is sent to the attackers. The sites may then request a one-time code, allowing the criminals to bypass multifactor authentication.
The campaign particularly targets advertising-account administrators, media buyers and agency employees. A compromised account can expose payment information, enable unauthorized advertising or lock legitimate users out.
Researchers have also linked the infrastructure to fraudulent Google Ads refund and employment pages, with paid advertisements used to distribute malicious links. TechRadar has reported on the campaign.
Users should visit only official websites and integrations, check the actual address bar and avoid entering a one-time code unless they initiated the login. Anyone who may have submitted credentials should change the password immediately, end active sessions and review administrators, permissions, payment details and recent account activity.
Source: Vesti Baku