DarkReading Researcher Tricks ChatGPT Into Building Undetectable Steganography Malware

[DarkReading]

Researcher Tricks ChatGPT Into Building Undetectable Steganography Malware

Using only ChatGPT prompts, a Forcepoint researcher convinced the AI to create malware for finding and exfiltrating specific documents, despite its directive to refuse malicious requests.

https://www.darkreading.com/attacks-brea...hy-malware