Skip to content
Submit incident
AII-1001-81999
Under investigationA regulator, court or police has opened a formal inquiry and reached no finding yet.Supporting sources are pending verification.

Microsoft sues alleged cybercrime network bypassing AI safety guardrails

Microsoft sued members of an international cybercrime network accused of stealing Azure OpenAI account credentials and selling access to malicious groups who used the system to generate non-consensual intimate images and other harmful content. The company identified four named individuals from Iran, Hong Kong, Vietnam and the UK, and prepared criminal referrals to law enforcement.

27 February 2025 Β· Storm-2139 Β· Security and cyberattacks
27 February 2025
AII-1001-81999
Quick facts
Category
Security and cyberattacks
Country
United States
Date
27 February 2025
Drafted by
Raidu Incident Detection

What happened

Microsoft identified members of a global cybercrime network called Storm-2139 who scraped customer logins from public sources and used them to access Azure OpenAI and other generative AI services.

The hackers bypassed safety guardrails on the AI tools and sold access to other malicious groups, providing instructions on how to create harmful content including non-consensual intimate images of celebrities and sexually explicit material.

Members identified by Microsoft include Arian Yadegarnia from Iran, Ricky Yuen from Hong Kong, PhΓ‘t PhΓΉng TαΊ₯n from Vietnam, and Alan Krysiak from the United Kingdom, with additional actors in Illinois and Florida.

Reported impact

The criminal group used the compromised accounts to generate thousands of images that violated Microsoft and OpenAI safety guidelines.

The harmful content included non-consensual intimate images of celebrities and other public figures, as well as sexually explicit material.

Index analysis

The attackers exploited publicly available stolen credentials to gain unauthorized access to legitimate Azure OpenAI accounts rather than directly attacking the systems.

Governance controls

Microsoft and OpenAI have technological controls in place to block the creation of harmful content, but the malicious groups gained access by using legitimate stolen customer credentials.

Microsoft is preparing criminal referrals to US and foreign law enforcement agencies.

Classification

Category
Security and cyberattacks
Industry
Digital security
Country
United States
Harm status
Actual harm
Intent
Intentional
Timing
Post-deployment
Harm types
Economic or property, Human or fundamental rights
Harmed
Consumers, General public
Failure modes
Jailbreak, Unauthorised scraping
Consequences
Litigation
AI systems
Azure OpenAI

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident