Microsoft names developers behind illicit AI tools used in celebrity deepfake scheme

Four foreign and two U.S. developers unlawfully accessed generative AI services, reconfigured them to allow the creation of harmful content such as celebrity deepfakes and then resold access to the tools, Microsoft said Thursday in a legal filing.

Users created “non-consensual intimate images of celebrities and other sexually explicit content” with the modified AI tools, including Microsoft’s Azure OpenAI services, the tech giant said in a blog post about its amended civil litigation complaint. The lawsuit was filed in December in a Virginia federal court and was unsealed in January. 

Microsoft did not name the celebrities out of concerns for their privacy. The company also said it “excluded synthetic imagery and prompts from our filings to prevent the further circulation of harmful content.”

The developers of the malicious AI tools are part of a “global cybercrime network” that Microsoft tracks as Storm-2139, the blog post said. 

The two U.S. individuals are based in Illinois and Florida, Microsoft said, but it withheld their names because of pending criminal investigations. 

The four foreign developers, the company said, are Arian Yadegarnia, aka “Fiz,” of Iran; Alan Krysiak, aka “Drago,” of the United Kingdom; Ricky Yuen, aka “cg-dot,” of Hong Kong; and Phát Phùng Tấn, aka “Asakuri,” of Vietnam.

Microsoft said it is preparing criminal referrals to law enforcement agencies in the U.S. and overseas.

Storm-2139’s access to the AI services was through “exploited exposed customer credentials scraped from public sources,” Microsoft said.

After Microsoft’s initial filing, the court issued a temporary restraining order and preliminary injunction that enabled the company to seize a website connected to Storm-2139. Microsoft said the disruption enabled its investigation to go deeper.

“The seizure of this website and subsequent unsealing of the legal filings in January generated an immediate reaction from actors, in some cases causing group members to turn on and point fingers at one another,” said the blog post, written by Steven Masada, assistant general counsel of Microsoft’s Digital Crimes Unit.

As chatter about the lawsuit increased, participants in the group’s communications channels also doxed Microsoft lawyers, “posting their names, personal information, and in some instances photographs,” the company said. The doxing backfired, though, and some suspected members of Storm-2139 emailed Microsoft, “attempting to cast blame on other members of the operation.” 

The six individuals mentioned in the blog post are among 10 “John Does” listed in the original complaint, Microsoft said.