Anthropic, a leading AI research firm, has released Claude Fable 5, the first publicly accessible version of its most powerful Mythos model series. The launch brings state-of-the-art AI capabilities to the public but comes with significant safety restrictions and a new data policy that could set a precedent for the industry.
Quick Facts
- Fable 5 is the first public version of the advanced Mythos model.
- The model blocks responses in high-risk areas like cybersecurity.
- A new mandatory 30-day data retention policy is now required.
A Cautious Rollout
Anthropic is making its new flagship model available through its Claude API and consumption-based Enterprise plans. Initially limited to a select group of partners over cybersecurity concerns, the technology is now being rolled out more broadly.
Subscription plan users will get temporary access. Through June 22, Fable 5 is included at no extra cost in Pro, Max, Team, and seat-based Enterprise plans. However, starting June 23, the model will be removed from these plans and will require usage credits, though Anthropic aims to restore it as a standard feature later.
Alongside the public release, the company is also deploying an updated version, Mythos 5, to the pre-approved organizations that already have access to the advanced model series.
The Price of Power and Safety
The release of Fable 5 follows recent warnings from Anthropic about the rapid pace of AI development, specifically the risk of systems achieving recursive self-improvement (RSI) without human oversight.
Reflecting these concerns, Anthropic has implemented hard safety limits. Fable 5 will block responses and defer to the older Claude Opus 4.8 model when prompted with queries related to high-risk fields like cybersecurity, biology, and chemistry. The company stated that it stress-tested the model against jailbreak attempts with both internal and external red-teaming organizations, which reportedly failed to find any universal exploits.
In a significant policy shift, access to Fable 5 and Mythos 5 now requires a 30-day data retention agreement on all traffic, overriding any previous zero-retention contracts with enterprise clients. Anthropic clarified the data will not be used for model training but solely to “defend against complex and novel attacks” and reduce false positives.
Impressive Performance, High Cost
Despite the restrictions, Fable 5 demonstrates formidable capabilities. Analytics company Hex reported that Fable 5 was the first model to score 90% on its benchmark for complex analytical tasks. Other platforms praised its performance, with Base44 noting its ability to “one-shot full apps,” and Genspark stating it beat all other models in its evaluations for UI design and game coding.
This performance comes at a premium. Fable 5 and Mythos 5 are priced at $10 per million input tokens and $50 per million output tokens—double the cost of Opus 4.8. The high price point may act as its own deterrent for widespread use, especially as enterprises become more critical of soaring AI operational costs.
However, some users believe the value justifies the expense. Shopping rewards platform Rakuten commented, “At the highest effort, Fable reflects on and validates its own work. For us, that’s what makes highly autonomous operations possible — the extra thinking pays for itself.”
Why This Matters for MENA
For the rapidly growing MENA tech ecosystem, Anthropic’s launch presents both opportunities and critical considerations. As governments and enterprises across the UAE, Saudi Arabia, and beyond accelerate their AI adoption, access to a powerful new model like Fable 5 is significant. Its noted strengths in software engineering and complex analytics are directly relevant to the region’s burgeoning fintech, logistics, and digital transformation sectors.
However, the mandatory 30-day data retention policy is a major sticking point. MENA-based organizations operate under increasingly strict data sovereignty and residency regulations, and a policy requiring traffic to be stored, even for safety, will demand careful legal and compliance reviews before adoption. Furthermore, the high cost will force regional startups and VCs to conduct a rigorous ROI analysis, weighing its advanced capabilities against the significant budget impact compared to alternatives from OpenAI, Google, or emerging regional players.
About Anthropic
Anthropic is an AI safety and research company dedicated to building reliable, interpretable, and steerable AI systems. The company focuses on research that advances the safety and capabilities of large-scale AI models to ensure they are beneficial for humanity.
Source: TechCrunch
