Antivirus vendors and security startups keep shipping AI features that promise to read malware the way a seasoned analyst would. The results inside security teams tell a quieter story. A new paper argues that static analysis of software, the job of deciding whether a program is malicious by examining its contents on disk, remains one of the hardest places to make generative AI work. The scale of the problem explains much of the difficulty. Standard … More →
The post A single malware file can outweigh an entire AI dataset appeared first on Help Net Security.
GPT-Red beat human red teamers on a prompt injection test
GPT-Red is an automated red-teaming model that OpenAI trains to find prompt injection weaknesses. It works the way a human red-teamer does. It sends a
