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

Microsoft has successfully led an industry-wide effort to dismantle 'EvilTokens,' a subscription-based cybercrime platform that utilized AI-driven chatbots to compromise 12,000 Microsoft accounts. Operating via Telegram, the service charged users a $1,500 entry fee followed by a $500 monthly subscription. The platform streamlined the process of email account compromise by automating inbox analysis, identifying high-value targets, and drafting sophisticated phishing emails. The integrated AI chatbot played a critical role by analyzing victim communications to identify trusted relationships and payment authorizations, allowing attackers to impersonate contacts and facilitate fraudulent fund transfers. By disrupting this service, Microsoft has curtailed a significant threat vector that enabled cybercriminals to execute complex fraud schemes in minutes rather than days. This operation highlights the growing trend of malicious actors leveraging generative AI to enhance the efficiency and success rate of large-scale social engineering attacks.