What happened
On 12 June 2026, Google said it had filed legal action against an alleged Chinese cybercrime network it calls **Outsider Enterprise**. According to Google, the group used AI platforms — including Google's own Gemini — to create fake websites and scam messages at scale.
The fake sites allegedly impersonated telecom providers, financial institutions, government agencies, and retailers, and people were lured to them through malicious text messages and ads. Google says hundreds of thousands of victims were affected, with losses reported in the millions of dollars.
To convey the scale, the reports cite figures such as roughly 9,000 fake websites, around one million fraudulent URLs, about 55,000 spam texts flagged by Android users, and some 2.5 million messages containing links to fraudulent sites over a two-week period. These are figures Google and the reporting attribute to the operation; the legal claims are still allegations.
Why it matters
The newsworthy part is not that scams exist — it is the role of AI. An LLM can write fluent copy, translate it, and generate page after page of believable-looking content in seconds. Work that once needed a team of people can now be done by a small group pointing an AI agent at the task.
For a beginner, the practical takeaway is a mindset shift. We are used to spotting scams by their rough edges — broken grammar, ugly layouts, obvious typos. AI removes those tells. A message can read perfectly, match a brand's tone, and still be a trap. So polish is no longer evidence of legitimacy. What still gives a scam away is where it actually sends you and what it asks you to do.
What to do next
A short, repeatable checklist beats trying to judge each message on its writing:
- **Check the sender and the domain.** Look at the real address and the link destination, not just the display name. Scammers use look-alike domains (extra words, swapped letters, odd endings).
- **Do not log in through links in texts or emails.** If a message says there is a problem with your bank or account, open the app or type the official address yourself.
- **Turn on stronger sign-in.** Use a passkey or two-factor authentication where it is offered, so a stolen password alone is not enough.
- **Report suspicious texts** through your phone's spam-reporting option. Those reports are part of how networks like this get traced and taken down.
- **Treat AI-grade polish as no proof at all.** Clean writing and a professional-looking page are easy to fake now. Verify the source independently before you click, pay, or share anything.
This briefing summarizes a public, dated report and links to its primary sources. The lawsuit's claims are allegations that have not been decided in court; the practical safety advice is general and applies regardless of the case's outcome.