Skip to content
Briefings are running a touch slower this week while we rebuild the foundations.See roadmap
AI: Jobs, Power & Money
10APR

Anthropic withholds Mythos from public release

2 min read
16:54UTC

Anthropic's most capable model scored 83.1% on vulnerability reproduction but will not be released publicly, going instead to twelve partners through a $100 million restricted programme.

← Back to The model they won't releaseJump to analysis ↓
EconomicDeveloping
Key takeaway

Anthropic set the precedent for withholding a frontier model from public release over systemic risk.

Anthropic released Claude Mythos Preview exclusively to twelve partner organisations through Project Glasswing on 8 April 2026, backed by $100 million in model usage credits 1. The model autonomously identified thousands of zero-day vulnerabilities across every major operating system and browser, including a 27-year-old OpenBSD flaw that had survived five million automated tests. On the CyberGym benchmark it scored 83.1% on vulnerability reproduction, compared with 66.6% for Anthropic's previous top model.

Anthropic has explicitly stated it will not release Mythos to the public. The twelve Glasswing partners include AWS, Apple, Google, Microsoft, CrowdStrike, Palo Alto Networks, and JPMorgan. Goldman Sachs, another partner, published displacement research the same week showing AI substitutes 25,000 jobs per month , placing these institutions on both sides of the AI capability and labour displacement story.

A Tom's Hardware review challenged the marketing: the "thousands" claim rested on only 198 manual reviews, and many flagged flaws were in outdated software 2. The Bessent-Powell emergency meeting suggests federal regulators took the risk seriously regardless.

Deep Analysis

In plain English

Every major AI model so far has been made available to the public, either free or via subscription. Anthropic has broken that pattern. Claude Mythos Preview is restricted to twelve partner organisations, including tech giants and financial institutions. The model can automatically find previously unknown security flaws in software at a scale that has never been seen from an AI system. Anthropic's position is that the capability is powerful enough that releasing it widely would create unacceptable risk, the same software that makes it useful to defenders could be used by attackers. So it is being distributed under a controlled programme called Project Glasswing, with $100 million in subsidised usage credits. A technical review by Tom's Hardware found that some of the specific claims were overstated, but the underlying capability gap the model represents is real.

Deep Analysis
Root Causes

Anthropic's capability assessment rests on the CyberGym benchmark jump from 66.6% to 83.1%, a 16.5-percentage-point improvement in autonomous vulnerability reproduction. Anthropic's own research on 'observed exposure' shows computer programmers face 75% task coverage from Claude-class models; Mythos's security capabilities represent the first instance where the coverage figure has operational implications beyond individual productivity.

The restriction decision reflects Anthropic's founding premise that AI safety and capability development must remain linked. Project Glasswing's $100 million credit allocation is structured as a subsidy for defensive deployment, not a commercial launch, which is itself novel for a frontier model.

What could happen next?
  • Precedent

    The first frontier AI model explicitly withheld from public release establishes capability-gating as a legitimate deployment option for safety-constrained AI systems.

    Medium term · 0.85
  • Risk

    The twelve Glasswing partners, which include both defensive (CrowdStrike, Palo Alto) and dual-use (AWS, Google, Microsoft) organisations, may deploy Mythos capabilities in ways beyond Anthropic's stated defensive intent.

    Short term · 0.62
  • Opportunity

    Security professionals who can interpret and direct AI-identified vulnerability data at scale represent a new premium-tier role the model creates even as it automates routine scanning.

    Medium term · 0.71
Sources:Anthropic
Mentions:Goldman Sachs →Carnegie Endowment for International Peace →Prima →Project Glasswing →Anthropic →Claude Mythos Preview →Amazon →Apple →Google →Microsoft →CrowdStrike →Palo Alto Networks →CyberGym →RUSI →JPMorgan Chase →
First Reported In

Update #5 · The model they won't release

Anthropic· 10 Apr 2026
Read original →
Causes and effects
This Event
Anthropic withholds Mythos from public release
The first frontier AI model deliberately withheld from public deployment establishes a precedent for capability-gated release in the industry.
Led to
Anthropic drops ASL, expands Glasswing partners
Anthropic's 7 April system card and Glasswing expansion directly extended the Glasswing programme launched on 8 April with the original twelve partners.
Occurred 7 Apr 2026
Read story →
Dimon: JPMorgan displaced workers from AI
JPMorgan is one of the original twelve Glasswing partners, creating the contradiction of a bank attending emergency Treasury meetings about AI risk while internally displacing staff through AI deployment.
Occurred 24 Feb 2026
Read story →
Different Perspectives
Directors Guild of America
Directors Guild of America
The DGA opened AMPTP talks on 12 May seeking AI training-use royalties that SAG-AFTRA and the WGA both settled without winning. France's SACD and European creative unions watch the DGA outcome as the US template for their own pending AI-training royalty negotiations with streaming platforms.
German IG Metall and European trade unions
German IG Metall and European trade unions
German unions led by IG Metall have pushed for binding co-determination rights on AI deployment since 2024; the Digital Omnibus literacy-duty weakening directly undercuts their model, which depends on a statutory information floor before works councils can challenge AI systems affecting members.
Chinese Ministry of Human Resources (MOHRSS)
Chinese Ministry of Human Resources (MOHRSS)
China's MOHRSS recognised 42 new AI occupations in April 2026 while Hangzhou courts upheld bans on AI-driven dismissal without retraining under the Labour Contract Law. Beijing's regulatory posture contrasts directly with Colorado's retreat: Chinese courts are adding employment liability for AI-driven redundancy while US courts remove state-level AI worker protection.
UK workers and Bank of England
UK workers and Bank of England
The ONS May 2026 bulletin showed payrolled employment down 210,000 year on year with no AI-specific breakdown, while the Bank of England's stress scenario used 500,000 additional unemployed as its AI-displacement worst case. UK workers are approaching that threshold through a dataset that cannot name its own cause.
India's IT sector workforce and NASSCOM
India's IT sector workforce and NASSCOM
NASSCOM's FY2026 data shows India's sector at 5.9 million while entry-level hiring fell 20 to 25%. GCC expansion by JPMorgan, Goldman Sachs and Apple benefits mid-career workers while closing the graduate entry pathway, replicating the under-25 displacement the NY Fed documented in US AI-exposed occupations.
European Parliament and Council (Digital Omnibus)
European Parliament and Council (Digital Omnibus)
The Digital Omnibus trilogue concession on AI-literacy duties reflects the Draghi report's argument that compliance overhead suppresses EU AI adoption. The Council traded the binding literacy mechanism for employer flexibility, leaving the December 2027 high-risk employment deadline without the worker-facing transparency layer Parliament had built around it.