Microsoft Busts Hackers Selling Illegal Azure AI Access

Microsoft's digital crimes unit identified and publicly named individuals linked to an LLMjacking campaign (Storm-2139) that harvested exposed API keys to sell unauthorized access to Azure AI services and manipulate models to produce illicit content; Microsoft filed lawsuits and seized infrastructure. The report details attacker methods (credential scraping, resale of access, bypassing model safeguards), real-world impacts (illegal imagery generation, resale on illicit marketplaces, doxxing), and recommends mitigations such as least-privilege access, vaulting API keys, and strong authentication.