Hacker's New Method: Creating Invisible Backdoor Malware Using OpenAI API

The Microsoft Security Team warned about a new type of malware called SesameOp that uses the OpenAI Assistants API for attacks. The software disguises legitimate cloud services as a covert command and control channel, allowing attackers to remain hidden in victims' systems. This attack method, first discovered in July 2025, highlights new risks of malicious use of cloud services.