Anthropic Eyes $950B Valuation as Its Secret Hacking AI Takes Over

📖 6 min read

Anthropic is in talks to raise between $30 billion and $50 billion at a valuation of up to $950 billion – a figure that would make it worth more than rival OpenAI and put it within striking distance of a trillion-dollar company. The round, first reported by The New York Times and Bloomberg on May 12, 2026, follows a $380 billion valuation just three months ago in February. That’s a 2.5x jump in 90 days.

The timing is no accident. Fueling investor appetite is Anthropic’s new flagship model, Claude Mythos – a cybersecurity-focused AI so capable at finding software vulnerabilities that Anthropic says it’s too dangerous to release to the general public. Yet despite that warning, the NSA is already using it. So are banks. So is half of Big Tech. Here’s the full picture.

The Numbers

Metric	Value
Fundraise target	$30B – $50B
Target valuation	Up to $950B
Previous valuation (Feb 2026)	$380B
Valuation increase in 3 months	+150% (2.5x)
OpenAI valuation (March 2026)	$852B
Anthropic annual revenue run rate	$30B
Expected revenue growth rate	~80% per year
Google investment commitment (April 2026)	$40B
Amazon investment commitment	$25B
Mythos usage credits committed	$100M
Donations to open-source security orgs	$4M

For context: the entire AI sector raised a record $297 billion in Q1 2026 alone. Anthropic is trying to capture a meaningful chunk of the next wave.

What Is Claude Mythos?

Mythos is Anthropic’s most advanced model to date – and it hasn’t been released to the public. Anthropic’s reasoning is blunt: the model is too good at offensive cyberattacks to let anyone use it freely.

Here’s what Anthropic has disclosed about Mythos’ capabilities:

• Found thousands of high-severity vulnerabilities across production software
• Identified weaknesses in every major operating system and web browser
• Can detect decades-old bugs that human security teams missed entirely
• Surpasses “all but the most skilled humans” at finding and exploiting software vulnerabilities, according to Anthropic

In response, Anthropic launched Project Glasswing – a defensive security coalition designed to get Mythos’ capabilities pointed at defense rather than offense. The partners list reads like a who’s-who of tech and finance: Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. Over 40 additional organizations have also been granted early access to scan their systems.

Anthropic is backing this with $100 million in free usage credits for Mythos Preview participants, plus $4 million in direct donations to open-source security organizations.

The Pentagon Paradox

Here’s where it gets strange. The Pentagon – specifically the Department of Defense under CTO Emil Michael – designated Anthropic as a “supply chain risk” after the two sides failed to agree on how Claude models could be used in military contexts. That’s the kind of label applied to vendors considered a potential threat to government system integrity.

Anthropic’s response: sue the Trump administration in March 2026 to overturn the blacklist.

But while the legal battle plays out, the NSA – which operates under the DoD – is reportedly using Mythos Preview anyway, according to Axios sources. The Pentagon’s own tech chief acknowledged Mythos is a “separate issue” from the blacklist. Translation: the government’s cybersecurity needs are overriding its political objections.

Anthropic is also briefing the Department of Homeland Security on Mythos’ capabilities. The model is simultaneously too dangerous for the public, restricted from regular DoD use, and essential enough that intelligence agencies won’t stop using it. That tension is, itself, a data point about how powerful Mythos actually is.

Why Investors Are Betting Nearly $1 Trillion on Anthropic

Three things make the $950 billion valuation less absurd than it sounds on first read.

1. Revenue is real and growing fast. CEO Dario Amodei said in early May that Anthropic has hit a $30 billion annual revenue run rate, with growth tracking at roughly 80% per year. That’s not projected revenue – it’s the annualized current pace. At $950 billion, investors are pricing Anthropic at roughly 32x forward revenue assuming 80% growth. High, but not unprecedented in high-growth enterprise software.

2. Infrastructure coverage is unmatched. Claude is the only frontier AI model available across all three major cloud platforms – AWS, Google Cloud, and Microsoft Azure. That gives it distribution no other model has. For enterprise buyers, that matters enormously. It means you can run Claude wherever your data already lives.

3. The big tech backstop is enormous. Google has committed up to $40 billion in Anthropic. Amazon has pledged up to $25 billion. These aren’t charitable donations – they’re strategic infrastructure bets. Cloud providers need a competitive AI model to counter OpenAI/Microsoft’s tight relationship. Anthropic is filling that gap, and both Google and Amazon are deeply incentivized to keep funding its growth.

Who’s Actually Competing With Anthropic?

Company	Valuation	Latest Model	Revenue Run Rate
Anthropic	$950B (target)	Claude Mythos (restricted)	$30B
OpenAI	$852B (March 2026)	GPT-5.5	Not disclosed
Google DeepMind	(part of Alphabet, $2T+)	Gemini (next gen pending)	Embedded in Google revenue
Meta AI	(part of Meta, $1.7T+)	Llama open-source	Open-source, no direct revenue

At $950 billion, Anthropic would surpass OpenAI in valuation despite being smaller in terms of brand recognition and consumer reach. The market is betting that enterprise penetration and cloud-platform access beats consumer mindshare.

The Risks Worth Watching

This isn’t a one-sided story. A few real concerns:

• The Pentagon lawsuit is unresolved. If Anthropic loses – or the legal battle drags on – it could restrict a significant potential government revenue stream, especially in defense and intelligence contracts.
• The Mythos delay has costs. By keeping its most powerful model restricted, Anthropic forfeits commercial revenue from the model until it can figure out safe deployment at scale. OpenAI’s approach has generally been faster commercialization even with known risks.
• $950 billion is still “talks.” The NYT notes explicitly that “the talks could fall apart.” Venture rounds at this scale have collapsed before. A completed deal is different from a term sheet rumor.
• Concentration risk. Google and Amazon together represent a huge chunk of Anthropic’s capital base. If either partnership sours – or either cloud provider develops a competitive model – Anthropic’s distribution advantage shrinks fast.

What This Means for You

If you’re using Claude today – for work, coding, writing, research – this round doesn’t change your experience directly. But it signals two things:

First, Anthropic isn’t going anywhere. A $950 billion valuation backed by Google and Amazon means the lights stay on. Claude isn’t a startup gamble anymore; it’s infrastructure.

Second, Mythos is coming. The restricted preview period won’t last forever. When it does open up – even in a constrained commercial form – it will represent a genuine step-change in AI capability, particularly for security-sensitive industries: finance, healthcare, government, and critical infrastructure.

If you’re in any of those fields and aren’t already testing Claude’s enterprise offerings, the window to get ahead of this before it’s mainstream is narrowing.

BetOnAI Verdict

Signal strength: High. The Anthropic story isn’t hype – it’s backed by $65 billion in committed capital from two of the largest companies on earth, real revenue growing at 80%, and a model powerful enough that intelligence agencies are using it against their own government’s orders. The $950 billion valuation is a bet that enterprise AI has a clear winner-take-most dynamic, and right now Anthropic is positioned better than anyone except possibly OpenAI.

The wildcards – the Pentagon lawsuit, Mythos’ delayed commercialization, and the deal not being closed yet – are real. But the direction of travel is unmistakable. Anthropic went from startup to near-trillion-dollar company in under four years. The next four years will determine whether it can actually earn that valuation or whether it joins the long list of tech unicorns that got ahead of their fundamentals.

For now, the smart money – literally – is betting yes.

---

Sources:

• New York Times – Anthropic in Talks to Raise Funding at a $950 Billion Valuation
• Anthropic – Project Glasswing announcement
• Axios – NSA using Anthropic’s Mythos despite Pentagon blacklist
• CNBC – Pentagon tech chief says Anthropic is still blacklisted, but Mythos is a separate issue
• New York Times – Anthropic CEO on $30B revenue run rate
• TechCrunch – NSA reportedly using Anthropic’s Mythos