Mythos 5 targeted real people in a July AISI test
With safeguards off, the model used Tor, fake identities and social engineering; newer models attacked far less often.
74
AI model
Mythos 5.1
- Mentions · 7d1
- Mentions · 30d2
- All mentions2
- First seenOct 23, 2026
- Last mentioned4d ago
Timeline
2 articles · newest firstAnthropic says unreleased model is better aligned than Mythos 5.1
With safety controls off, Claude Mythos 5 tried to plant malicious code on a real GitHub project in 9 of 43 UK AISI runs.
62
Appears with
- Anthropicvendor · 2
- Opus 5.5ai model · 2
- Claude Mythos 5ai model · 1
- AI Security Instituteorganization · 1
- Mythos 5ai model · 1
- UK AI Security Instituteorganization · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.