Capabilities · Actor
Claude Fable 5 / Claude Mythos 5 / Claude Opus 4.8 (Anthropic)
Anthropic Claude model family including redeployed Fable 5 in the mid-2026 product lineup.
Forecast horizon
Cyber safeguard details and jailbreak framework may set precedent for future model deployment governance.
Connections
Connections · 7
How this node ties into the rest of the map, and the evidence behind each link.
from · publishes 4/10
Anthropic Anthropic developed and deployed Fable 5 and Mythos 5 as frontier models.
+4 growth to · applies to 3/10
CogManip: Multi-Turn LLM Manipulation Benchmark CogManip evaluates frontier models including GPT-5.4 and DeepSeek-V3.2; Claude-class models are in scope for such manipulation benchmarks.
+3 growth from · constrains 3/10
Trump AI Executive Order / US Export Control Directive on Frontier AI (2026) US government export control directive suspended all access to Fable 5 and Mythos 5.
+3 growth from · applies to 3/10
HackAgent Automated Red-Teaming Framework HackAgent red-teaming framework was applied to evaluate adversarial robustness of Fable 5 and Opus 4.8.
+3 growth from · applies to 3/10
Industry-Wide Jailbreak Severity Scoring Framework The jailbreak severity framework was proposed alongside the redeployment of Fable 5 and its cyber safeguards.
+3 growth from · applies to 3/10
Claude for Physical AI (UST Deployment) UST is applying Claude models to physical AI use cases, extending Anthropic's model family into robotics/industrial contexts.
+3 growth from · supports 3/10
Claude Opus 5 (Anthropic) Opus 5 extends the Claude flagship model line for agentic and coding use.
+3 growthSignal sources
Signal sources
Dated facts from primary sources in this direction.
Task horizon doubling Mar 2025
The length of software tasks AI agents can do autonomously at 50% reliability has doubled about every 7 months — and since 2024 closer to every ~3 months.
METR → Benchmarks saturating Apr 2025
In one year scores rose by 18.8, 48.9 and 67.3 points on MMMU, GPQA and SWE-bench; real-world software solve rate jumped from 4.4% to 71.7%.
Stanford HAI — AI Index 2025 → Autonomous coding 2025–2026
On SWE-bench Verified (500 real GitHub issues), autonomous coding agents reached ~80–86% by late 2025, up from under 50% in early 2025.
Epoch AI →