Microsoft researchers bullish on AI security agent even though it let 74% of malware slip through

Project Ire promises to use LLMs to detect whether code is malicious or benign
Microsoft has rolled out an autonomous AI agent that it claims can detect malware without human assistance.…