ChatGPT blindly trusts browser content, turning the page into a payload
A Russia-linked threat group known as GREYVIBE has exploited a vulnerability in ChatGPT, using the AI tool to craft phishing lures and payloads. Researchers discovered that ChatGPT blindly trusts browser content, allowing malicious web pages to be transformed into phishing tools. This tactic was employed throughout a campaign targeting Ukrainian military and government entities. AI
IMPACT Exploitation of ChatGPT's trust in web content highlights security risks for AI-powered tools and their users.