Issue 2026-09-09 · Industry · 安全
Hackers stealing Claude tokens, Anthropic investigates

On August 4, consultant Grant De Swardt found token usage climbing on his Claude Max 20x account despite inactivity. Anthropic suspended his paid account, invalidated sessions and server-side Claude Code tokens, refunded £44.49 for remaining subscription time, and said a compromised session key was used to mint unauthorized OAuth tokens.
Read original ↗- User Says Anthropic Wrongly Tagged Them for Training Competitor Model
- Anthropic study: harmless-looking data can train models to sabotage oversight
- Anthropic model finds Linux kernel bug enabling remote root access
- HN users question sudden Opus 5.5 quality drop
- Johns Hopkins cancer mapping tech picked for Anthropic AI for Science