On June 12, Google filed a civil lawsuit in a New York federal court against a China-based cybercrime network called Outsider Enterprise. The accusation is unusual: the group used Google's own Gemini AI to mass-produce phishing websites and scam text messages that drained bank accounts and cryptocurrency wallets across dozens of countries.
The operation ran like a factory floor. Members coordinated through Telegram channels, sharing ready-made prompts for Gemini that could generate convincing fake telecom and banking pages in minutes. Those pages were fed directly into the group's toolkit and deployed at scale — resulting in more than 8,000 fraudulent websites, 2.5 million scam text messages, and, according to FBI estimates, 3.87 million stolen payment card numbers. Total losses since July 2023 are put at roughly $1.9 billion.
Cryptocurrency users were a specific target. Some phishing pages mimicked exchange login screens and wallet interfaces to capture seed phrases and passwords. Once entered, recovery is essentially impossible since blockchain transactions cannot be reversed. In just two weeks ending June 1, Google logged 55,000 suspicious message complaints in Google Messages alone.
What sets this lawsuit apart is the legal theory behind it. Rather than handing evidence to prosecutors, Google is suing the group directly in civil court — arguing that unauthorized, harmful use of an AI platform is its own cause of action. "We intend to permanently dismantle this criminal organization," the company said. If that argument holds up, it could give AI companies a new legal tool against groups that turn their own models against users.
The Outsider Enterprise case maps out what AI-enabled crime now looks like structurally. The group ran as a phishing-as-a-service platform: affiliates got Telegram access, received templates and AI prompts, and launched attacks without writing a single line of code themselves. Generative AI lowers the skill floor for everyone — including criminals. How widely this model has spread to other platforms is the question the industry has yet to face squarely.
