Ars Technica
Microsoft disrupts AI-assisted platform that compromised 12,000 accounts

Microsoft disrupted EvilTokens, a subscription-based scam platform that used an AI chatbot to compromise 12,000 Microsoft accounts in a few months. The service, launched via a Telegram channel in February, charged $1,500 upfront and $500 monthly, and guided criminals through inbox analysis, target selection, and email drafting to deceive employees into transferring funds. The AI chatbot identified trusted relationships, payment authorizations, and sensitive responsibilities to recommend fraud strategies.