Hackers are exploiting Claude subscriptions to consume tokens without user activity, as first reported by a user last month. Anthropic has since issued a warning to its user base about the unauthorized token usage.
Background
Anthropic's Claude is a popular AI assistant with a token-based subscription model. Token theft attacks on AI platforms are an emerging threat as AI services grow in popularity.
- Source
- TechCrunch
- Published
- Sep 9, 2026 at 05:10 AM
- Score
- 6.0 / 10