Phishing Platform Impersonates AI Chatbot Ad Portals to Steal Credentials and MFA Codes

Sophisticated Phishing Operation Uncovered

Cybersecurity researchers have detailed a new 'human-operated phishing platform' that creates fake advertising portals for popular AI chatbots. These include Google Gemini, Anthropic Claude, OpenAI ChatGPT, Perplexity, Meta Muse, and Manus. The deceptive sites claim to offer services like campaign optimization and ad account connections.

Browser-in-the-Browser Attack Method

The core of the attack involves the browser-in-the-browser (BitB) trick. When a user clicks a 'Connect' button on the phishing site, a fake browser window appears within the real browser. This fake window displays a spoofed address bar showing trusted origins, such as accounts.google.com or an Okta tenant, while the actual browser remains on the phishing domain. This method is used to capture login credentials and multi-factor authentication (MFA) codes.

Real-Time Credential Harvesting

Behind the user interface, the platform records every password attempt and fingerprints the victim's device. An operator then selects which MFA challenge the victim sees next. For example, a site like 'museads.ai' emerged shortly after Meta launched its AI agent, Muse, and claimed to be an 'AI ads manager.' Clicking its 'Connect' button initiates a BitB attack to steal credentials for Google, Meta, TikTok, and Okta.

Operator Control and Data Exfiltration

Victim device information is sent to the attacker, and operator commands are exchanged based on the login workflow. With the captured account credentials, the attacker attempts to log into the victim's account in real-time. Each fake brand portal offers a tailored pitch, such as ChatGPT promising Google Ads briefs or Gemini offering MCC support, to entice users.