# Anthropic's Fable

> **TL;DR.** Anthropic's Fable is the community and media shorthand for [Claude Fable 5](https://www.

- **Category:** AI / Frontier Models / Anthropic
- **Stage:** nascent
- **Age:** 3 days
- **Origin date:** 2026-06-09
- **First detected:** 2026-06-12
- **Canonical URL:** https://earlyterms.com/term/anthropic-s-fable
- **Sources:** 7 primary URLs

## Definition

Anthropic's Fable is the community and media shorthand for [Claude Fable 5](https://www.anthropic.com/claude/fable) — Anthropic's first publicly available Mythos-class model, launched June 9, 2026 for long-horizon agentic work, multi-day software engineering sessions, and complex multi-step reasoning at $50/M output tokens.

The informal phrase crystallized not at launch but in the 48-hour controversy that followed: a hidden guardrail silently degraded responses for AI researchers and cybersecurity professionals without visible notification. Anthropic apologized on June 11 and made the safeguard transparent, but the "Anthropic's Fable" framing in TechCrunch, Fortune, and Engadget headlines had already fixed the association between the model and its guardrail debate.

## Analogy

Think of it as the model's public persona — the version shaped by the story the press told, not Anthropic's naming.

## Why it's emerging now

Claude Fable 5 launched June 9, 2026 as the first public Mythos-class model — but within 24 hours a hidden guardrail silently poisoned AI researchers' outputs, triggering a "secret sabotage" wave across tech media. Anthropic reversed course June 11, but "Anthropic's Fable" as a controversy-coded phrase had already displaced the branded name in the news cycle.

## Related terms

- *alias:* Claude Fable 5
- *alias:* claude-fable-5
- *parent:* claude-mythos
- *parent:* claude-mythos-preview
- *related:* claude-opus-4-8
- *related:* agentic-coding
- *related:* managed-agents
- *related:* agent-harness
- *related:* long-running-agents
- *child:* silent sabotage mode

## Sources

1. [Anthropic — Claude Fable landing page (canonical product page)](https://www.anthropic.com/claude/fable)
2. [TechCrunch — Cybersecurity researchers aren't happy about the guardrails on Anthropic's Fable](https://techcrunch.com/2026/06/10/cybersecurity-researchers-arent-happy-about-the-guardrails-on-anthropics-fable/)
3. [Fortune — Anthropic accused of 'secret sabotage' as Fable silently limits AI researchers](https://fortune.com/2026/06/10/anthropic-accu-claude-fable-5-limits-capabilities-ai-researchers-developers/)
4. [Gizmodo — Anthropic apologizes for Fable 5 guardrail and will change it](https://gizmodo.com/anthropic-apologizes-for-one-of-the-guardrails-on-its-fable-5-model-and-will-change-it-2000770365)
5. [Hacker News — Cybersecurity researchers guardrails thread (584 pts / 517 comments)](https://news.ycombinator.com/item?id=48478969)
6. [Engadget — Anthropic's Fable AI brings Mythos capabilities to regular users](https://www.engadget.com/2190934/anthropic-fable-ai-brings-the-capabilities-of-its-unreleased-mythos-model-to-regular-users/)
7. [Anthropic — Claude Fable 5 and Claude Mythos 5 launch announcement](https://www.anthropic.com/news/claude-fable-5-mythos-5)

---
_Generated by EarlyTerms · https://earlyterms.com/term/anthropic-s-fable_