Microsoft has just struck hard in its fight against the misuse of artificial intelligence. The company has revealed the identities of four individuals accused of illegally exploiting its Azure OpenAI service to generate deepfakes for sexual nature, including celebrities.
Microsoft has revealed that individuals from Iran, the United Kingdom, Hong Kong and Vietnam are believed to be key members of an international criminal network called Storm-2139 by Microsoft. According to the company, this group allegedly used API keys stolen from several Microsoft customers to access the Azure OpenAI service and bypass safeguards put in place to prevent the creation of harmful content.
Microsoft's Digital Crimes Unit initially filed a complaint against ten "John Doe" individuals (unidentified individuals) for violations of U.S. law and the Terms of Service for Generative AI Services. This complaint has just been amended to specifically name and identify four individuals: Arian Yadegarnia (aka "Fiz"), Alan Krysiak (aka "Drago"), Ricky Yuen (aka "cg-dot"), and Phát Phùng Tấn (aka "Asakuri").
Also read – Copilot betrays Microsoft, AI reveals how to hack Windows 11
A sophisticated criminal network
Microsoft's investigation revealed a complex organization within Storm-2139. The group is believed to consist of creators who develop the malicious tools to bypass the protections of AI services, vendors who modify and resell these tools at various prices, and end users who use them to generate illicit synthetic content.
The defendants allegedly not only exploited illegally obtained customer credentials, but also modified the capabilities of the AI services to resell them to other malicious actors. They allegedly provided detailed instructions on how to generate harmful content, including non-consensual intimate images of celebrities and other sexually explicit content.
As a reminder, Microsoft obtained a court order allowing it to seize a website crucial to the criminal group's operations, thus disrupting its ability to market its illegal services. Steven Masada, deputy general counsel for Microsoft's Digital Crimes Unit, stressed the importance of exposing these individuals and shining a light on their malicious activities to set a precedent in the fight against the misuse of AI technologies.
0 Comments