VERDICT —OUT OF SCOPE
Root cause is private-key / signer compromise — the on-chain contract behaved exactly as written. No pre-deployment source audit or bytecode review reaches the key-custody perimeter; this is operational-security territory (HSM/MPC hygiene, key rotation, hot-wallet isolation). Bytecode would show nothing wrong.
▰ METHOD
PRIVATE KEY
PRIVATE-KEY
Root cause
Root-cause analysis not yet published. The incident description below contains all currently available signal — review the attack transaction directly for definitive forensics.
Forensic narrative
The official X account of AI startup Anthropic, backed by Amazon, appears to have been compromised, posting an unknown token contract address related to AI Agents. Attack method (per SlowMist): Account Compromise. Reported loss: -.
Primary source
https://x.com/AnthropicAI/status/1869139895400399183 ↗Sourced from
slowmist
Technical record
- chain
- —
- protocol
- Anthropic
- bug_class
- private-key
- date_occurred
- 2024-12-18
- loss_usd
- —
- source_id
- sm:anthropic::2024-12-18
Related — same bug class· private-key