Fake ChatGPT, Claude and Gemini Ads Use Browser-in-the-Browser Phishing to Steal Accounts

A human-operated phishing campaign impersonates AI advertising tools (Muse Ads, ChatGPT-, Gemini-, and other AI-branded fronts) to trick advertising professionals into connecting accounts; malicious pages use a Browser-in-the-Browser fake auth UI and Socket.IO APIs to collect credentials, MFA codes, and device attributes, with researchers observing hundreds of victim submissions and multiple ad-themed domains linked as IOCs. The platform uses Next.js frontends and Railway backends, enables operators to request/validate MFA challenges, and targets manager accounts that can expose multiple clients—leading to account takeover, campaign hijacking, and prolonged recovery times; recommended mitigations include reviewing sessions, removing unfamiliar users/partners, pausing unauthorized campaigns, and auditing connected assets.