Claude Mythos Preview
Anthropic's Claude Mythos Preview is no longer just a model — it's a government-endorsed cyber weapon. Developed by Anthropic, it cleared all UK cyberattack simulators while doubling METR time horizons at 80% success. Mozilla, XBOW, Project Glasswing, and even the National Security Agency all use it. The UK AI Safety Institute both endorses and regulates it, alongside the AI Security Institute. But the competition is tightening: GPT-5.5 tied it in enterprise cyber tests, and Codex 5.3, GPT-5.3, and Claude Opus 4.6 all vie for the same space. With only 6 mentions in 30 days, buzz is low for a model this consequential. The question: can Mythos maintain its security lead as rivals close the gap?
- ·Cleared all UK cyberattack simulators; doubled METR time horizon at 80% success.
- ·Adopted by Mozilla, NSA, XBOW, and Project Glasswing; endorsed by UK AI Safety Institute.
- ·Competes directly with Codex 5.3, GPT-5.3, Claude Opus 4.6, and GPT-3.5.
- ·GPT-5.5 tied Mythos in enterprise cyber tests, signaling competitive parity.
Signal Radar
Five-axis snapshot of this entity's footprint
Mentions × Lab Attention
Weekly mentions (solid) and average article relevance (dotted)
Timeline
11- Research MilestoneMay 14, 2026
First AI model to clear all UK AISI cyberattack simulations
View source - Research MilestoneMay 9, 2026
Early snapshot achieves more than 2x time horizon of next best model on METR benchmark
View source - Research MilestoneMay 2, 2026
Claude Mythos Preview scored 68.6% on AISI expert CTF tasks
View source- score:
- 68.6%
- task type:
- expert CTF
- Research MilestoneMay 2, 2026
Claude Mythos Preview fully solved TLO enterprise network simulation in 3 of 10 attempts
View source- simulation:
- TLO
- success rate:
- 3/10
- Research MilestoneApr 15, 2026
First AI model to complete full AISI cybersecurity evaluation
View source - Research MilestoneApr 15, 2026
Leaked benchmark results suggest it 'destroys every other model' including GPT-5, Claude 4 Opus, and Gemini Ultra 2.0
View source - Research MilestoneApr 14, 2026
Achieved 73% success rate on expert-level CTF challenges and completed full 32-step network attack simulation
View source- ctf success rate:
- 73%
- network attack success:
- 3 of 10 attempts
- Regulatory ActionApr 14, 2026
Evaluated by UK AI Safety Institute for autonomous cyber attack capabilities.
- evaluator:
- UK AI Safety Institute
- test type:
- capture-the-flag and network simulation
- Research MilestoneApr 14, 2026
First AI model documented to autonomously complete a full, multi-step cyber attack simulation in UK safety tests.
- success rate:
- 73%
- simulation success:
- 3 of 10 attempts
- steps:
- 32-step network takeover
- Research MilestoneApr 12, 2026
Scored 83.1% on the CyberGym benchmark for vulnerability discovery.
View source- score:
- 83.1%
- benchmark:
- CyberGym
- Product LaunchJan 1, 2026
Anthropic released Claude Mythos Preview model, used by Mozilla for security bug triage and patching.
View source
Relationships
17Competes With
Uses
Developed
Regulated
Endorsed
Recent Articles
5Claude Mythos Clears All UK Cyberattack Simulators, Doubling Speed Revised
+Claude Mythos Preview became the first AI model to clear all UK AISI cyberattack simulations, forcing the agency to double its capability-doubling est
100 relevanceOpenAI Launches Daybreak Cyber Initiative to Rival Anthropic's Glasswing
+OpenAI launched Daybreak, a cybersecurity initiative using GPT-5.5 and Codex Security, to rival Anthropic's Glasswing project.
92 relevanceClaude Mythos Preview Doubles METR Time Horizon at 80% Success
+Claude Mythos Preview snapshot achieves 2x METR time horizon over next best model at 80% success rate, per Anthropic. Absolute numbers undisclosed.
89 relevanceClaude Mythos Helped Firefox Fix More Bugs in April Than 15 Prior Months Combined
+Firefox fixed more security bugs in April 2026 than 15 prior months combined, using Anthropic's Claude Mythos Preview model for triage and patching.
86 relevanceGPT-5.5 Ties Claude Mythos in Enterprise Cyber Attack Tests, AISI Finds
+UK AISI finds GPT-5.5 matches Claude Mythos on full enterprise network attack simulation, scoring 71.4% on expert tasks vs 68.6%.
100 relevance
Predictions
No predictions linked to this entity.
AI Discoveries
2- observationactiveApr 15, 2026
Velocity spike: Claude Mythos Preview
Claude Mythos Preview (ai_model) surged from 1 to 3 mentions in 3 days (velocity_spike).
80% confidence - observationactiveApr 10, 2026
Velocity spike: Claude Mythos Preview
Claude Mythos Preview (ai_model) surged from 0 to 7 mentions in 3 days (new_surge).
80% confidence
Sentiment History
| Week | Avg Sentiment | Mentions |
|---|---|---|
| 2026-W15 | 0.16 | 9 |
| 2026-W16 | 0.43 | 3 |
| 2026-W17 | 0.35 | 2 |
| 2026-W18 | 0.30 | 1 |
| 2026-W19 | 0.65 | 2 |
| 2026-W20 | 0.55 | 2 |