Attackers Abuse ChatGPT Custom GPTs to Deliver RAT via ClickFix Lures

According to Huntress researchers, threat actors have begun abusing ChatGPT Custom GPTs to masquerade as legitimate product offerings. These malicious GPTs are designed to direct unsuspecting victims to external websites that employ ClickFix lures. ClickFix is a social engineering technique that tricks users into executing commands, typically copied from a webpage, under the guise of fixing an issue or verifying they are not a robot. Instead of resolving a problem, executing the command delivers a Remote Access Trojan (RAT) onto the victim's system.

Huntress stated this activity marks the abuse of yet another feature in trusted artificial intelligence (AI) platforms. Prior campaigns have weaponized shared links and other functionalities, and this latest tactic leverages the credibility associated with the ChatGPT brand to lower victim defenses. The attackers customize GPTs to appear as helpful tools or product-specific assistants, prompting users to follow links that lead to malware delivery infrastructure.

OpenAI has been notified of the activity. The company is likely reviewing and removing the malicious Custom GPTs identified by Huntress. Security experts recommend that users verify the legitimacy of any links generated by AI platforms and avoid executing unknown commands from untrusted sources. The incident highlights the evolving threat landscape where trusted AI tools are co-opted for malware distribution.