Hackers are stealing Claude tokens from subscribers
A Claude user noticed unexplained token usage while inactive, prompting Anthropic to warn users that hackers may be accessing accounts and consuming resources without permission.
MAIN POINTS
- A Claude user saw token consumption despite not actively using the account.
- The incident suggested unauthorized access rather than normal user activity.
- Anthropic responded by warning users about possible hacker involvement.
- The warning highlights account security risks for AI service users.
TAKEAWAYS
- Unexpected token usage can signal compromised account access.
- Users should monitor AI account activity for anomalies.
- Security warnings from providers may indicate broader attack patterns.
- Protecting credentials is important to prevent unauthorized resource use.