Anthropic
AI safety company building frontier foundation models including Claude; created the Mythos cybersecurity model withheld from public release due to its ability to find zero-day exploits in critical infrastructure.
Metrics · from transcript
$100M compute credit fund for defensive security hardening