Microsoft Sues Over 10 John Does in AI Hacking Scheme

Microsoft has filed a lawsuit against three individuals and seven of their customers, all identified as John Does, for running a “hacking-as-a-service” operation. This service allowed the creation of harmful content by bypassing Microsoft’s AI safety measures. The defendants developed tools to circumvent these safety protocols and compromised legitimate customer accounts. They operated from July to September last year, selling access through a now-closed site. The service used a proxy server and undocumented APIs to mimic legitimate requests to Microsoft’s Azure servers. Microsoft’s complaint, filed in Virginia’s Eastern District federal court, aims to disrupt this sophisticated cybercrime. The company did not specify how the accounts were compromised but noted that hackers often search code repositories for inadvertently published API keys.

Source: arstechnica.com