@ai_and_law · Post #487 · 20/01/2025 08:04
🇺🇸Microsoft Takes Legal Action Against AI Abuse Scheme Microsoft has launched two lawsuits targeting an elaborate scheme to bypass safety guardrails in its generative AI systems. The first lawsuit is aimed at three individuals accused of running a "hacking-as-a-service" operation. These defendants allegedly created tools to exploit undocumented APIs, compromised legitimate customer accounts, and built a fee-based platform for generating harmful and illicit content. The second lawsuit targets ten individuals who were customers of this service. According to the complaints, the defendants’ platform provided detailed instructions on bypassing Microsoft’s restrictions, enabling the creation of prohibited content, including materials promoting harm or discrimination. Microsoft alleges the platform used stolen API keys and proxy servers to mimic legitimate requests, undermining the integrity of its Azure OpenAI systems. Both lawsuits, filed in Virginia, invoke the Computer Fraud and Abuse Act, the Digital Millennium Copyright Act, and other legal provisions. These cases underscore the ongoing challenges in securing generative AI platforms against abuse while maintaining their accessibility for ethical uses. #AIRegulation#Cybersecurity#EthicsInAI#Microsoft#AIAbuse