Home / Technology / Anthropic releases advanced AI with strong safeguards
Anthropic releases advanced AI with strong safeguards
9 Jun
Summary
- Claude Fable 5, an advanced AI model, has been released by Anthropic.
- Significant safeguards are in place for Fable 5 to prevent misuse.
- Mythos 5, the same model as Fable 5, will have safeguards lifted for some users.
Anthropic has introduced Claude Fable 5, an AI model that sits above its Opus class in capability. This "Mythos-class" model comes with significant safeguards, leading to its distinct name. For queries concerning potentially damaging capabilities, such as in cybersecurity, Fable 5 will defer to the next most capable model, Claude Opus 4.8, ensuring responsible deployment.
Anthropic is also continuing its Project Glasswing initiative. Through this, a select group of cyberdefenders and infrastructure providers will gain access to Claude Mythos 5. This model is identical to Fable 5 but will have certain safeguards lifted in specific areas, highlighting its advanced cybersecurity features. The company plans to broaden access to Mythos 5 via a trusted program soon.
The release of Fable 5 follows Anthropic's recent confidential S-1 filing, signaling preparations for an IPO. This move occurs shortly after competitor OpenAI also confidentially filed its S-1. Concerns about the advanced capabilities of AI models, particularly regarding software and cybersecurity vulnerabilities, had previously caused market and governmental unease.
