TL;DR: Anthropic warned that infostealer malware is stealing Claude login sessions to hijack accounts and drain usage.
Summary: Anthropic issued a security warning about infostealer malware that is actively stealing Claude login sessions, giving attackers the ability to hijack accounts and consume allotted usage. The warning underscores that AI assistant accounts are now a direct target for credential-stealing campaigns.
Why it matters: AI builders should enforce session monitoring, MFA, and least-privilege API keys to prevent account takeover and usage abuse in their own Claude integrations.
Source: rss