Anthropic released Claude Fable 5 on Tuesday, giving public users access to its most powerful class of AI models for the first time.
Key Points:
- Claude Fable 5 is built on the same tier of technology as Anthropic’s restricted Mythos models.
- The company says broad safeguards will limit risky queries in cyber, biology and chemistry.
- The release follows new White House efforts to expand voluntary AI safety testing.
Anthropic Fable
The company announced the model on Jun. 9, saying Fable 5 outperforms its earlier public systems across software engineering, knowledge work, vision, scientific research and other benchmarks.
According to Jared Perlo of NBC News, Fable 5 belongs to the same family as Anthropic’s Mythos models, which drew concern earlier this year because of their ability to find and exploit cyber vulnerabilities.
Anthropic said Fable 5 uses Mythos-level technology but is safe for general use because it has added limits and safeguards. The company said sensitive prompts tied to cybersecurity, biology and chemistry will be routed to Opus 4.8, its previous top public model.
Also Read: Crypto Hack Fears Grow Around Anthropic’s Possible Claude Fable Release
Claude Safety
AI researchers have warned that stronger models could help attackers target banks, power grids and other critical infrastructure. Others have raised concerns that advanced systems could assist bioweapon design.
Anthropic said it is taking a conservative approach at launch, while still making other Mythos-level features available to Pro, Max, Team and seat-based Enterprise users at no extra cost until Jun. 22.
The release came one week after Donald Trump signed an executive order creating a voluntary process for AI companies to share models with the government for safety testing before public release.
Read Next: Kalshi Lists Chainlink Perp As LINK ETF Assets Cross $101M