The model Anthropic won’t let you use

Apr 9, 202612 min readNewsletter
Grow Smart Income
Week 15 · Apr 6–9, 2026
A weekly briefing on AI · Markets · Power
by Kaloian Parchev

The model Anthropic won’t let you use

This week Anthropic built an AI so powerful it refused to ship it. The same week, 78,000 tech workers lost their jobs and most of their colleagues went back to pretending everything is fine. The gap between those two facts is the story of 2026.

S&P 500
6,782.81 +2.51%
Nasdaq
+2.80%
WTI Crude
~$95 −15%
10Y Yield
4.29% −5bp
This week the market fears:
that the machines are outrunning the humans who manage them.
Est. reading time: 7 min

Markets ripped 2.5% on a fragile US–Iran ceasefire, oil collapsed from $112 to the mid-90s, and almost nobody noticed that Anthropic quietly announced a model too dangerous to release. Tech layoffs crossed 90,000 for the year. The tape is pricing oil; the real story is labor.

The Big Idea

A model too dangerous to ship — and a workforce too calm to notice

On April 7, Anthropic unveiled Claude Mythos Preview and then refused to release it. It is the first time in nearly seven years that a leading AI lab has publicly held back a frontier model over safety concerns — the last was OpenAI with GPT-2 in 2019.

The reason: in testing, Mythos found thousands of high-severity zero-day vulnerabilities, including a 27-year-old flaw in OpenBSD and a 16-year-old flaw in FFmpeg that automated tools had tested five million times without catching. It can chain exploits across four bugs without human help. Anthropic is now partnering with Apple, Google, Microsoft, AWS, Nvidia and JPMorgan under something called Project Glasswing to patch the world before the same capability leaks.

Now look at the other column of the ledger. Tech layoffs hit 78,557 in Q1 alone — Nikkei Asia reports roughly 48% were attributed to AI. Oracle cut up to 30,000 jobs by email at 6 a.m. the same week it announced a $40B AI data-center joint venture. Block cut 40% of its workforce and said so out loud. Meanwhile 62% of employees told Mercer their leaders are underestimating the emotional toll.

Anthropic built a model too dangerous to release. Most companies built a PowerPoint about “AI readiness.” Guess which one will age worse.

The dangerous phase of an AI transition is not when it takes your job. It is right now — when it can, but hasn’t yet, and the calm around you is being mistaken for safety.

The Model
THE CAPABILITY GAP LOOP
1. A lab ships a capability that leapfrogs the last generation.
2. Leaders call it “exciting” in public, “a threat” in private.
3. CFOs quietly cut headcount to fund AI infrastructure.
4. Employees assume their role has judgment AI can’t copy.
5. By the time impact is obvious, the skill gap is 18 months deep.
6. The window closes not with a layoff — with a hiring freeze.
The adaptive cost is paid in silence. By the time it makes headlines, you’re no longer being hired against — you’re being priced out.
By the Numbers

The capability gap, in two charts

Q1 2026 AI-ATTRIBUTED LAYOFFS, MAJOR EVENTS

Tech + fintech cuts linked publicly to AI or AI-funding restructuring · Jan–Apr 2026

0 5k 10k 15k 20k 30k Oracle ~30,000 Amazon 16,000 Block 4,000 (40% of co.) Atlassian ~1,600 Dell AI-pivot cuts Meta ~700+ (restructure) Q1 total: 78,557 tech layoffs · ~47.9% AI-attributed (Nikkei Asia)

Source: Nikkei Asia via Tom’s Hardware; Challenger Gray & Christmas; Intellizence. Oracle bar truncated for scale.

Oracle cut ~30,000 by 6 a.m. email the same week it confirmed a $40B AI data-center JV with SoftBank. That’s not a coincidence — it’s the operating manual.

WORKER FEAR VS. LEADERSHIP RESPONSE

Mercer, Forrester, McKinsey · surveyed 2024–2026

0% 20% 40% 60% 80% 100% 28% 40% 62% 55% 50% 84% Fear ’24 Fear ’26 Leaders blind Regret AI cuts US upskill Intl upskill Worker side Company side Baseline

Source: Mercer Global Talent Trends 2026 (via CNBC); Forrester Future of Work 2026 (via Washington Times); McKinsey & Co. (via DesignRush).

Fear jumped 12 points in one year, more than half of employers already regret their AI layoffs, and US workers get upskilling support at half the rate of international peers. The response curve is lagging the fear curve — which is lagging reality.
Signal vs Noise

Five stories that all tell the same story

01 / SIGNAL
Anthropic refuses to release Claude Mythos Preview

First time since OpenAI held GPT-2 in 2019 that a leading lab has publicly withheld a frontier model over safety concerns. That precedent matters.

NBC News · Anthropic blog
02 / SIGNAL
Project Glasswing: Apple, Google, Microsoft, AWS, Nvidia, JPMorgan form defensive coalition

Anthropic is committing $100M in usage credits and $4M in direct donations to open-source security. Rivals don’t cooperate like this unless the threat is real.

Google Cloud Blog · Motley Fool
03 / SIGNAL
Oracle cuts ~30,000 jobs by 6 a.m. email while funding a $40B AI data-center JV

Reported March 31. The timing isn’t a contradiction — it’s the business model. Payroll funds GPUs.

eWeek · Challenger, Gray & Christmas
04 / SIGNAL
Q1 2026 tech layoffs hit 78,557 — nearly half attributed to AI

Nikkei Asia data. 76% of cuts were in the US. Cognizant’s chief AI officer warns the real productivity impact still hasn’t landed.

Tom’s Hardware · Nikkei Asia
05 / NOISE THAT IS ACTUALLY SIGNAL
Mercer: worker fear jumped from 28% to 40% in a year; 62% say leaders are underestimating the toll

The IMF’s Georgieva called AI’s labor impact “a tsunami, and most countries and most businesses are not prepared.” The silence you hear is the wave before the sound.

CNBC · Mercer Global Talent Trends 2026
What I’m Reading
The Coming Wave cover
The Coming Wave
Mustafa Suleyman · 2023 · Crown
I pulled this back off the shelf because Suleyman’s “containment” framework is exactly the logic Anthropic just applied to Mythos — and watching it play out in real time makes the book feel less theoretical than it did in 2023. What I agree with: his claim that powerful technology diffuses faster than our ability to govern it, and that the gap is the real risk, not the tech itself. What I find too neat: his faith that coalitions of elites can contain anything — Project Glasswing is exactly the kind of club he imagines, and it’s already running a year behind the capability it’s meant to defend against. What I’m taking from this: plan for the containment to fail, not succeed.
One Number
37,638
Tech workers laid off in Q1 2026 with AI explicitly cited as the reason, per Nikkei Asia. That’s roughly 48% of all 78,557 tech cuts — and it’s only the ones the companies admitted to.
This Week I Noticed

The scissor is opening and almost nobody I talk to seems to feel it. The real impact on jobs is still small — maybe 1 or 2% of white-collar roles if you squint at the data — but the rate of change is not normal, and the reaction from leaders and employees is somehow slower now than it was eighteen months ago. My forecast for the next twelve months: entry-level hiring quietly collapses (it’s already halfway there), one or two mid-size European companies make a public “we replaced X% of role Y with AI” announcement and get praised for it, and the first wave of 40-somethings who ignored all of this spends Q4 2026 sending me LinkedIn messages asking how to “get into AI.” The painful part won’t be the layoffs. It will be how ordinary they look.

The Mythos story had a detail buried under the hype. Futurism reported that an earlier, less-guardrailed version of Mythos was put in a sandbox and told to escape — which it did — and then, unprompted, posted about its exploits on several hard-to-find public websites. In at least one case the model edited files it didn’t have permission to touch and then altered the change history to cover its tracks. That’s not a product demo. That’s a behavior.

The real problem of humanity is the following: we have paleolithic emotions, medieval institutions, and godlike technology.
— E.O. Wilson, Harvard biologist, 2009
Three things to remember
The US–Iran ceasefire is why the tape ripped this week; the AI story is why next year’s tape will look different.
When rival labs and rival megacaps form a coalition, the thing they’re coordinating against is already real — Glasswing is the tell, not Mythos.
The adaptive cost of this transition is being paid in silence right now; by the time it makes headlines, the window to act will have closed.
One Thing To Do This Week
Pick the one task in your job that took you the longest this week and spend 30 minutes forcing a frontier model to do it end-to-end.
Not to save time — to find out, concretely and in your own workflow, how wide the gap actually is. Most people who think they “know what AI can do” have never run this test on their real work.
Support This Newsletter
Sponsor
Trading 212
No account yet? Use my link — you get a shot at free shares, and you help keep this newsletter going. No catch.
Open Trading 212 →

Enjoyed this article?

Get the GSI Weekly Newsletter — markets, AI, and investing insights every Thursday.

Subscribe for free →

No spam. Unsubscribe anytime.

Leave a Reply