A critical vulnerability in ChatGPT could be used by cybercrooks to gain unauthorized access to sensitive information.
The flaw allows attackers to inject malicious URLs into input parameters, leading to unintended requests on their behalf.
The bug is classified as of medium severity but its Exploit Prediction Scoring System score has significantly increased to 55.36%.
Steps to address the threat include monitoring ChatGPT usage, implementing data validation techniques, updating software and systems, and educating employees about risks.