Microsoft Takes Legal Action Against AI “Hacking as a Service” Scheme

Microsoft alleges three individuals ran a hacking-as-a-service platform that circumvented Azure OpenAI safety guardrails by relaying requests through a proxy service using undocumented APIs and compromised API keys to mimic legitimate traffic, and has initiated legal action. The piece outlines the scheme’s method of bypass and account compromise, framing it as abuse of generative AI services rather than a specific vulnerability disclosure or breach.