Anthropic faces major security breach as hackers target AI computing power

Anthropic recently disclosed a large-scale security breach affecting its Claude platform, with numerous user accounts being forcibly logged out and payment methods removed. The incident was attributed to malware such as Vidar, Lumma, and AMOS, which infiltrated users' systems through deceptive means like cracked software. These malware strains stole sensitive data, including browser passwords and session cookies, allowing hackers to bypass two-factor authentication and access user accounts without passwords. Hackers are now exploiting stolen Claude accounts to monetize AI computing power and API quotas, selling access to unsuspecting users or leveraging proxy platforms to charge by token usage. Anthropic took drastic measures, including deleting stored credit card information and revoking access to infected accounts. Users are advised to log out of all devices, clear browser data, and avoid non-official software to prevent infections. The breach highlights the growing threat of AI-related cybercrime in the current digital landscape.