Anthropic is moving its most powerful and carefully restricted AI model, Claude Mythos, from a tightly controlled partner program into commercial products for paying enterprise customers. Evidence emerged this week when the model identifier "claude-mythos-1-preview" briefly appeared in the live Claude interface, and Anthropic confirmed on May 22 that it plans a general release once safeguards are ready.

Why Has Anthropic Kept Mythos Locked Away?

Mythos surfaced publicly in March when a misconfigured Anthropic content management system exposed internal documents describing it as "by far the most powerful AI model we've ever developed". The model excels at finding security vulnerabilities at superhuman speed, which is why Anthropic treated it as a national security concern. The UK National Health Service (NHS) walled off GitHub repositories because of Mythos capabilities. The White House drafted executive order language around frontier model vetting partly because of Mythos. The Pentagon excluded Anthropic from classified networks partly because of the model's potential.

Instead of a public release, Anthropic launched Project Glasswing on April 7, a restricted program giving access to roughly 40 partner organizations. These partners included:

• Cloud and Tech Giants: AWS, Microsoft, Google, and Apple
• Hardware and Security Leaders: NVIDIA, CrowdStrike, and the Linux Foundation
• Financial Services: JPMorgan Chase

The results justified the caution. Glasswing partners and Anthropic's own teams have found more than 10,000 high or critical-severity vulnerabilities in essential software. Anthropic's public coordinated disclosure dashboard, updated May 22, lists 1,596 specific disclosures across 281 open-source projects, with 97 patched so far. This is not a chatbot; it is a vulnerability-finding engine that has already shifted the global security baseline.

How Is Anthropic Planning to Release Mythos Safely?

The commercialization strategy follows a deliberate pattern. Anthropic is building two specific access points for Mythos 1: Claude Code, the developer terminal being transformed into an agent platform, and Claude Security, a new enterprise product with a redesigned dashboard. The Security dashboard includes seven-day and thirty-day historical charts, deeper vulnerability triage, and a unified vulnerability view.

Anthropic has already laid groundwork through its Opus 4.7 model rollout, which included built-in cybersecurity safeguards that automatically detect and block high-risk offensive use. Multiple researchers and reporters have flagged this as the stepping stone toward a Mythos-class release, with the architecture for safely shipping frontier capabilities being built in pieces, in public.

Steps to Understanding Anthropic's Dual-Use AI Strategy

• Leak Forces Transparency: An accidental exposure of internal documents revealed Mythos existed and what it could do, forcing Anthropic to acknowledge the model publicly rather than keep it secret.
• Restricted Partner Program Proves Value: Project Glasswing demonstrated that defenders needed Mythos before attackers could access it, justifying the controlled release to 40 organizations.
• Enterprise Product Captures Revenue: Claude Code and Claude Security will monetize Mythos capabilities through subscription-based access, generating revenue while maintaining safeguards.
• General Release Eventually Follows: Anthropic has committed to broader access once safeguards mature, establishing the playbook the entire industry will likely follow for sensitive capabilities.

This pattern represents how dual-use AI gets commercialized in practice. Anyone evaluating AI vendors should expect every frontier lab to copy this structure for capabilities too sensitive for an open API.

What Does This Mean for Enterprise Security Teams?

The cybersecurity stack is being rewritten in real time. If Mythos 1 lands in Claude Security with the new dashboard, every enterprise security team faces a choice: continue running existing vulnerability-management stacks, or integrate a model that has already found more critical bugs in major software than most security firms find in a year. The procurement conversation in serious Chief Information Security Officer (CISO) offices is about to change fundamentally.

Anthropic's $30 billion funding round, which closed this week, reinforces this trajectory. The deal exceeded $30 billion at a pre-money valuation above $900 billion, making Anthropic the world's most valuable private AI company, surpassing OpenAI's $852 billion valuation from March. The valuation jump from $380 billion in February to over $900 billion in May is among the fastest at this scale in venture history. Multiple sources expect this to be Anthropic's final private round before an October public offering.

The funding close and the Mythos commercialization story are not separate events. The investor pitch for a $900 billion valuation requires showing what Anthropic ships next that justifies it. Mythos 1 in Claude Code and Claude Security is that answer. Anthropic's projected Q2 revenue of $11 billion and first operating profit provide the financial foundation for the valuation.

The honest read is not that Mythos launches tomorrow. Rather, Anthropic is constructing the channel through which Mythos reaches paying customers, and it will be Claude Code and Claude Security first, not a free chat product. The model that scared governments is becoming a product, behind gates, with safeguards, but unmistakably wider than 40 partners.