Fake ChatGPT Models Spread via Google Ads Using ClickFix Malware Delivery Technique

Cybersecurity firm Huntress discovered a malicious campaign exploiting the legitimate ChatGPT domain to distribute a counterfeit GPT model called "Plus 5.6" through paid Google ads. The scam directs users to a fake CAPTCHA verification page that tricks them into executing a Windows command, installing a remote access trojan called "@input."

How the Scam Works

This sophisticated fraudulent campaign combines multiple deception techniques. Attackers invest in Google search ads to ensure their fake GPT model appears at the top of results when users search for ChatGPT.

Warning Signs and Red Flags

The most critical indicator is the presence of the fake model within the legitimate ChatGPT domain itself. When users select the model labeled "Plus 5.6," they encounter what appears to be a legitimate error message stating "The basic service has limited availability," which then redirects them to another page.

On that page, users see what closely resembles an official Cloudflare CAPTCHA verification screen. However, instead of typical CAPTCHA verification, the page instructs users to copy and paste a specific command into Windows Command Prompt.

Legitimate security verification processes never request such actions.

Dangers After Infection

Executing the command installs a remote access trojan (RAT) called "@input" on the system. The permissions this malware grants are extensive:

Screen monitoring and recording
Camera and microphone access
File searching and extraction
Installation of additional malware
Remote system control

Attackers can steal private information or use the compromised device for financial fraud without the victim's knowledge.

Prevention Strategies

Never execute commands requested by websites: Regardless of which website or page requests Windows Command Prompt input, never comply. This is a certain indicator of fraud.
Don't assume safety based on domain legitimacy: In this case, a fake model existed within the legitimate chatgpt.com domain. Even if the domain is authentic, not all content within it is secure.
Be cautious of paid search ads: Fraudulent sites can appear in Google search results despite being paid advertisements. Always verify the URL shown at the bottom of the ad and confirm it directs to the genuine website.
Verify through official channels: If uncertain about ChatGPT, avoid clicking search results. Instead, directly visit OpenAI's official website (openai.com) to verify information.
Install trusted security software: Deploy reliable malware protection software and keep it updated regularly.

Where to Report and Seek Help

If infection is suspected: Immediately isolate the device from the network and consult with a trusted cybersecurity professional
In the United States: Contact the FBI's Internet Crime Complaint Center (IC3) or the Cybersecurity and Infrastructure Security Agency (CISA)
International reporting: INTERPOL or your country's national cybercrime authority