Anthropic releases Fable 5 with safeguards as Mythos-class AI goes public

Anthropic's Claude Fable 5 brings Mythos-class intelligence to general users, but cybersecurity queries automatically downgrade to a less capable model.

Anthropic released Claude Fable 5, a Mythos-class AI model that downgrades cybersecurity and biological research queries to its older Opus 4.8 system, marking the first time the company has made its most powerful technology broadly available.

"We wanted to be able to provide this level of intelligence for general users in a safe manner," Dianne Penn, Anthropic's head of product management, research and labs, told The Wall Street Journal.

Fable 5 costs $10 per million input tokens and $50 per million output tokens, less than half the price of Claude Mythos Preview. The model outperforms all previous Claude systems across software engineering, vision, and knowledge-work benchmarks. Stripe used Fable 5 to complete a codebase-wide migration in a 50-million-line Ruby repository in a single day, a task the company estimated would have required more than two months of manual engineering effort.

The release signals that Anthropic believes its safety classifiers are robust enough to commercialize frontier AI without exposing users to its most dangerous capabilities. But early testing suggests the safeguards may be broader than advertised, potentially limiting the model's utility for cybersecurity professionals who represent a key customer segment.

How the safeguards work

Fable 5 uses AI-powered classifiers that detect requests related to cybersecurity, biology, chemistry, and model distillation. When triggered, the system routes the query to Claude Opus 4.8 instead of Fable 5. Anthropic says the fallback occurs in fewer than 5 percent of sessions, meaning most users interact with the full Mythos-class model during ordinary use.

The company describes the safeguards as intentionally conservative, prioritizing safety over convenience. Anthropic ran an external bug bounty that produced no universal jailbreaks in more than 1,000 hours of testing, though the UK AISI made progress toward one within a brief initial testing window.

Rob T. Lee, chief AI officer and chief of research at the SANS Institute, told CSO that his routine cybersecurity tasks involving incident response, detection, and basic forensic workflows were automatically routed from Fable 5 to Opus 4.8 during initial testing. If those observations hold under broader testing, it could indicate that Anthropic's classifiers are broadly identifying cybersecurity-related requests rather than distinguishing between benign and malicious activity.

Mythos 5 for cyber defenders

For a select group of users, Anthropic is also releasing Claude Mythos 5, the same underlying model as Fable 5 but with cyber safeguards removed. Through Project Glasswing, about 200 organizations including Verizon and Microsoft will gain access to the unrestricted version. Anthropic plans to gradually expand access through a broader trusted-access program developed in consultation with the US government.

The company says Mythos 5 possesses the strongest cybersecurity capabilities of any model currently available, including the ability to discover software vulnerabilities, assist with exploit development, and perform complex multi-stage cybersecurity tasks. Those capabilities are precisely what prompted Anthropic to restrict access to earlier versions of the technology.

What it means for security leaders

For CISOs and security teams, the announcement raises questions about how quickly organizations can adapt to increasingly capable AI systems. The challenge is no longer simply obtaining access to advanced models but integrating them into security operations in ways that produce measurable benefits.

Anthony Grieco, Cisco's senior vice president and chief security and trust officer, said organizations should focus on deploying powerful models effectively while maintaining strong security fundamentals. "The pace of frontier AI development is changing the security landscape in real-time, and defenders cannot afford to wait for the dust to settle," Grieco said in a statement.

At the same time, Grieco cautioned against viewing AI as a substitute for foundational security practices. "AI will raise the ceiling for what defenders can do, but security resilience remains the foundation that determines whether those gains translate into real protection," he said. Even as AI models accelerate software engineering, analysis, and security operations, organizations still need to execute on fundamentals such as patching, multifactor authentication, network segmentation, and zero trust architectures.

Anthropic shares, which do not trade publicly, have no direct market impact. But the release pressures rivals OpenAI and Google to match Mythos-class capabilities while maintaining comparable safety standards. OpenAI announced its own confidential S-1 filing on Monday, and SpaceX, which includes Elon Musk's xAI, is set to begin trading on the Nasdaq this week.

This article is for informational purposes only and does not constitute investment advice.