The 25% Unemployment Prepper: Inside Anthropic’s $350 Million Plan to Save You from the AI Apocalypse

 

 The 25% Unemployment Prepper: Inside Anthropic’s $350 Million Plan to Save You from the AI Apocalypse

**Subtitle:** *From 5% unemployment to the “unprecedented,” the maker of Claude just dropped a triage manual for the end of work. Here is the blueprint—and the math—for universal basic income, sovereign wealth funds, and why the company is taxing itself.*

**Reading Time:** 8 Minutes | **Category:** Technology & Economics

## Introduction: The “Intrinsic” Feature

Dario Amodei, the CEO of Anthropic, has a confession to make. It is not one that will win him friends in boardrooms. But it is one that he believes is too important to ignore.

“There is a decent possibility that, despite efforts to soften the blow, AI could cause significant enduring job loss,” Amodei wrote in a lengthy policy essay this week . “And this may be an intrinsic property of the technology and the way it broadly replicates human cognition” .

In other words, the job losses are not a bug. They are a feature.

For months, Amodei has been warning that AI could eliminate half of entry-level white-collar jobs within five years and push unemployment to 10% to 20% . This week, he unveiled a **$350 million plan** to prepare for the consequences .

The plan is not a single policy. It is a **triage manual**—a set of escalating responses calibrated to the severity of the labor market disruption. It lays out specific recommendations for three scenarios: a world with 5% unemployment, a world with 10% unemployment, and a world of “unprecedented” unemployment .

In the mildest scenario, Anthropic proposes expanding “pre-distributive capital accounts”—baby bonds seeded at birth—and incentivizing companies to retrain workers rather than lay them off . In the medium scenario, the priority is expanded unemployment insurance, sector-specific transition support, and basic-needs relief . In the worst-case scenario—the one that could approach 25% unemployment—Anthropic calls for permanent income replacement funded by new taxes on AI companies, higher capital gains taxes, or broad-based consumption levies .

“The key challenge in such a world won’t be incentivizing growth, but finding a way for everyone to share in the benefits,” Amodei wrote .

In this deep-dive, we will break down the three-tiered framework, analyze the “unprecedented” scenario that gets the most attention, and critique the feasibility of a plan proposed by the very company causing the disruption.

> **The Bottom Line Up Front:** Anthropic is not waiting for governments to act. It is writing the playbook itself. The $350 million commitment funds research and fellowships, not direct cash payments. But the framework is a roadmap for the post-work world—and a warning that the AI industry expects the transition to be brutal.

## Part 1: The “Intrinsic” Feature – Why Job Loss Isn’t a Bug

To understand Anthropic’s urgency, you have to understand Amodei’s argument about the nature of the technology.

### The “General Substitute” Thesis

Previous waves of automation targeted physical labor. Factory robots replaced assembly line workers. ATMs replaced bank tellers. In each case, new jobs were created to replace the old ones .

AI is different. Amodei argues that AI is a **“general substitute for human labor”** . It is not replacing a specific task. It is replacing cognition itself. And when a technology can replicate the core cognitive functions of the white-collar workforce, the jobs may not come back.

“This reframes one of the AI industry’s most uncomfortable questions,” Business Insider noted . “If AI systems are designed to perform more of the cognitive work humans do, then job losses may not simply be the result of bad corporate behavior or short-term adjustment. Amodei suggests they could be a structural consequence of successful AI development.”

### The 10% to 20% Warning

This is not the first time Amodei has sounded the alarm. Previously, he warned that AI could eliminate half of entry-level white-collar jobs within five years and push unemployment to 10% to 20% .

His latest essay is less about predicting a specific jobs apocalypse than spelling out what governments should do if enduring displacement arrives . The framework is a recognition that the “intrinsic” feature of AI may be fewer jobs—and that society needs a plan.

### The “Doom Loop” Data

Anthropic’s own internal data illustrates the pace of acceleration. As of May 2026, more than **80% of the code merged into Anthropic’s own codebase** was written by its AI assistant, Claude . The average Anthropic engineer now ships **eight times as much code** per quarter as they did before AI.

“The company’s internal data shows just how fast the technology is accelerating,” The Times of India reported . If Anthropic’s own engineers are being augmented to the point of 8x productivity, what happens to the rest of the software industry?

**The Human Touch:** For the software engineer reading this, the 80% statistic is personal. It is not an abstraction. It is the code review they did yesterday—or the code review they were passed over for because Claude did it faster.

| White-Collar Field | Estimated AI Displacement Risk (Amodei) |

| :--- | :--- |

| **Entry-Level Coding** | Very High |

| **Entry-Level Finance** | Very High |

| **Entry-Level Law** | Very High |

| **Customer Support** | Already Replaced |

| **Data Entry** | Already Replaced |

| **Mid-Level Management** | Moderate |

| **Senior Leadership** | Low |

*Sources:  *

## Part 2: The Three Tiers – A Triage Manual for the End of Work

Anthropic’s framework is calibrated to the severity of labor market disruption . It is not a one-size-fits-all solution. It is a set of escalating responses.

### Tier 1: The 5% Unemployment Scenario (Where We Are Now)

The current US unemployment rate is 4.3% . Anthropic’s “mild” scenario assumes a rise to roughly 5%—a level that is historically uncomfortable but not catastrophic.

**The Recommendations:**

- **Expand “pre-distributive capital accounts”:** Baby bonds seeded at birth, with expanded eligibility for young adults. Currently, these accounts can hold only index funds. Anthropic proposes allowing them to hold stakes in AI companies .

- **Incentivize retention:** Create tax credits and other incentives for companies that retain and retrain workers rather than laying them off .

- **Workforce training grants:** Fund programs to help displaced workers transition to new roles .

- **Wage insurance:** For workers who have to take lower-paying jobs due to AI displacement, provide insurance to cover the wage gap .

**The Subtext:** We are already in this scenario. The recommendations are designed to be implemented now, before the disruption accelerates.

### Tier 2: The 10% Unemployment Scenario (The Great Recession Level)

The last time US unemployment hit 10% was during the Great Recession of 2009 . Anthropic’s second scenario assumes a similar level of disruption.

**The Recommendations:**

- **Expand unemployment insurance:** Lengthen the duration and increase the amount of benefits .

- **Sector-specific transition support:** Targeted aid for industries hit hardest by AI displacement (e.g., call centers, data entry, entry-level legal and accounting) .

- **Basic-needs relief:** Cash assistance for food, housing, and healthcare .

- **Manage the pace of rollout:** If AI is a general substitute for human labor, policymakers should consider incentivizing firms to manage displacement gradually rather than all at once .

**The Subtext:** This is where the transition gets painful. The social safety net as we know it is not designed for double-digit unemployment. The framework calls for a temporary expansion of existing systems.

### Tier 3: The “Unprecedented” Scenario (25% Unemployment)

This is the scenario that gets the attention. Anthropic does not specify an exact number, but the historical reference point is the Great Depression, when US unemployment hit 25% in 1933 .

**The Recommendations:**

- **Income replacement for a large share of the workforce:** The policy challenge shifts from “temporary transition” to “permanent support” .

- **New sources of tax revenue:** The framework proposes several mechanisms :

  - *Taxes on AI companies:* Measured by tokens, compute, or revenue .

  - *Higher capital gains taxes:* Taxing the wealth generated by AI-driven asset appreciation .

  - *Broad-based consumption taxes:* A VAT or national sales tax .

- **New ways to share broadly:** The revenue would fund :

  - *Universal basic income:* A regular cash payment to all citizens .

  - *Sovereign wealth models:* A national investment fund owned by the public .

  - *Equity-sharing mechanisms:* Giving workers partial ownership in AI enterprises .

**The Subtext:** Anthropic is less certain about the right answers here. “This scenario is novel economic territory, so we’re less certain about the right answers here,” the company wrote . The $350 million commitment is designed, in part, to research exactly these mechanisms.

| Scenario | Unemployment | Primary Policy Response |

| :--- | :--- | :--- |

| **Tier 1** | ~5% | Capital accounts, retention incentives, training, wage insurance  |

| **Tier 2** | ~10% | Expanded UI, sector-specific aid, basic-needs relief  |

| **Tier 3** | Unprecedented (25%+) | UBI, sovereign wealth funds, equity-sharing, new taxes  |

## Part 3: The Funding Mechanism – Taxing the Machines

The most novel—and controversial—aspect of Anthropic’s framework is the funding mechanism for the worst-case scenario.

### The “Tax on AI” Proposal

Amodei has suggested that a universal basic income could be financed through taxes on “relevant companies” or by raising the capital gains tax . The specifics are not fleshed out, but the principle is clear: the companies that profit from AI displacement should pay for the social safety net.

“Potential revenue sources could include increasing the capital gains tax, broad-based consumption taxes, sector-specific levies on AI use (measured by tokens, compute, or revenue), and scalable ‘digital dividends’ funded by taxes on the digital sector,” the framework states .

### The “Sovereign Wealth” Model

An alternative mechanism is a **sovereign wealth fund**—a national investment fund owned by the public, funded by taxes on AI-driven productivity gains . The returns would be distributed as a dividend to all citizens.

This model has precedent. Alaska’s Permanent Fund distributes oil revenue to every resident. Norway’s sovereign wealth fund is the largest in the world. The difference is that the “resource” being taxed would be AI productivity, not oil .

### The “Equity-Sharing” Mechanism

A third option is **equity-sharing**—giving workers partial ownership in AI enterprises . This is distinct from UBI. It is not a handout. It is a stake.

“Equity-sharing mechanisms giving workers partial ownership in AI enterprises” could take the form of stock grants, profit-sharing, or broad-based employee ownership plans .

**The Human Touch:** For the worker facing displacement, the difference between UBI and equity-sharing is the difference between charity and ownership. One says “we will support you.” The other says “you still have a stake in the future.”

| Funding Mechanism | Mechanism | Political Viability | Example |

| :--- | :--- | :--- | :--- |

| **AI Company Tax** | Tax on tokens, compute, or revenue | Low (industry opposition) | Digital services tax |

| **Capital Gains Tax** | Tax on AI-driven asset appreciation | Moderate (class warfare) | Biden-era proposals |

| **Consumption Tax** | VAT or national sales tax | Low (regressive) | European model |

| **Sovereign Wealth Fund** | Public ownership of AI-driven returns | Low (novel) | Alaska Permanent Fund |

| **Equity-Sharing** | Worker ownership stakes | Moderate | ESOPs |

*Sources:  *

## Part 4: The Skeptics – “Fear-Mongering” or “Genuine Warning”?

Not everyone is applauding Anthropic’s initiative. The most prominent critic is **David Sacks**, the former White House AI Czar under President Trump .

### The “Hypocrisy” Charge

Sacks has been scathing. He accuses Anthropic of “fear-mongering” about AI’s dangers while racing ahead with development.

“Signs you might be trying to get your frontier AI lab nationalized,” Sacks wrote in response to Amodei’s essay . “You compare it to nukes… threaten half of white-collar jobs… warn recursive self-improvement could end humanity… then race ahead anyway. In other words, you want the government to save us from… you” .

Sacks also pointed out a specific hypocrisy: Anthropic has a job listing for a software engineer on its website with a salary of $570,000 . “So what Anthropic is saying is they’re still trying to hire software engineers at a very high wage... but somehow they think these jobs are going to be eliminated,” he said .

### The “Convenient” Timing

The $350 million commitment comes just as Anthropic is preparing for a hotly anticipated IPO . Some critics argue that the timing is not coincidental.

“The executives—who once highlighted AI’s disruptive effects—are now spending more time discussing how workers and society can benefit from the technology’s gains as they gear up for hotly anticipated IPOs,” Business Insider reported .

In other words, the “we’ll take care of the workers” messaging is good for business. It positions Anthropic as the responsible AI company, differentiating it from rivals like OpenAI.

### The “Intrinsic” Defense

Amodei has anticipated the criticism. His essay explicitly states that he is not “trying to be a ‘prophet of doom’” . He is trying to prepare.

“We are not seeking job displacement,” Anthropic wrote in the framework . “We are working to prevent or minimize it. Some amount of displacement, though we cannot say how much, may be an intrinsic consequence of the technology, and our responsibility is to prepare for it and respond to it.”

The question is whether preparation is a substitute for prevention—or an acknowledgment that prevention is impossible.

**The Human Touch:** For the software engineer who applied for the $570,000 job at Anthropic, the Sacks critique is personal. The company is warning that AI will eliminate entry-level coding jobs. And it is offering half a million dollars to senior engineers. The two realities are not contradictory—but they are revealing.

| Critic | Argument | Anthropic’s Defense |

| :--- | :--- | :--- |

| **David Sacks** | Fear-mongering; racing ahead; hypocrisy | Preparing for intrinsic consequence |

| **Industry Observers** | Convenient timing (IPO) | Long-standing position |

| **Economists** | Unclear policy details | Framework, not final plan |

## Part 5: The Path Forward – What Happens Now

The $350 million commitment is the first step. Here is what comes next.

### The Economic Futures Research Fund

The $200 million Economic Futures Research Fund will back research trials and “program evaluation” on public policies that Anthropic considers promising . This is not a direct cash payment program. It is a research initiative.

The goal is to generate evidence on what works—and what doesn’t—before the worst-case scenario arrives.

### The National Fellowship Program

The $150 million national fellowship program will help early-career professionals “extend the benefits of AI to communities across America” . The details are sparse, but the intent is to democratize access to AI tools.

### The Policy Engagement

Anthropic is actively engaging with policymakers. The framework is written for a US government audience . The company is also following the lead of its rival, OpenAI, whose CEO Sam Altman recently met with Senator Bernie Sanders to discuss a public wealth fund .

President Trump has indicated that he will meet with AI executives to discuss “giving back” to the public, telling reporters that if they do so, “the public will become very wealthy” .

### The Open Questions

The framework leaves many questions unanswered.

- **Who decides when the thresholds are crossed?** Is it the government? An independent body? Anthropic itself?

- **How are the new taxes enforced?** AI tokens cross borders. Compute can be rented anywhere. A national tax on AI may be unenforceable.

- **What about the transition?** Even if the policies are adopted, the transition will be painful. The framework does not address the human cost of the interim.

**The Human Touch:** For the worker who is laid off next month, the $350 million fund is cold comfort. The research will take years. The policies will take longer. The displacement is happening now.

## Frequently Asked Questions (FAQ)

**Q: What did Anthropic announce?**

A: Anthropic announced a **$350 million commitment** to study AI’s impact on jobs and the economy. The funding includes a $200 million Economic Futures Research Fund and a $150 million national fellowship program . The company also published a policy framework for how governments should respond to AI-driven labor displacement .

**Q: What are the three scenarios in the framework?**

A: The framework addresses three levels of labor market disruption: **5% unemployment**, **10% unemployment**, and an **“unprecedented” level** (potentially 25% or higher) . Each scenario has escalating policy responses .

**Q: Does Anthropic support universal basic income?**

A: In the worst-case scenario—unprecedented unemployment—the framework lists UBI as one of several possible mechanisms for income replacement, alongside sovereign wealth funds and equity-sharing . Amodei has suggested that UBI could be funded through taxes on AI companies or higher capital gains taxes .

**Q: Is Anthropic the only AI company talking about this?**

A: No. OpenAI recently announced that it wants the gains from AI to be “widely shared,” and CEO Sam Altman has explored proposals for a public wealth fund . President Trump has also indicated he will meet with AI executives to discuss “giving back” to the public .

**Q: Is the $350 million going directly to displaced workers?**

A: No. The funding is for research and fellowships, not direct cash payments . The research fund will back trials and evaluations of public policies. The fellowship program will help early-career professionals use AI to support communities.

**Q: Why is Anthropic doing this now?**

A: The announcement comes as Anthropic prepares for a hotly anticipated IPO . Critics argue the timing is convenient, positioning the company as responsible. Anthropic argues that the warnings are genuine and the preparation is urgent.

## Conclusion: The “Unprecedented” Preparation

We started this article with a number: 25%. That is the unemployment rate of the Great Depression—the historical precedent for Anthropic’s “unprecedented” scenario.

We end with a different number: **$350 million**. That is what Anthropic is spending to prepare.

The AI industry is acknowledging what many have feared: the disruption is not temporary. The job losses are not a bug. They may be an intrinsic feature of the technology.

**For the Worker:**

Do not wait for the government. Do not wait for Anthropic. The transition is happening now. The best defense is to build skills that AI cannot easily replicate: judgment, creativity, relationship management.

**For the Policymaker:**

The framework is a starting point. The questions of enforcement, timing, and transition are unanswered. The time to answer them is now—before the “unprecedented” scenario arrives.

**For the Citizen:**

The debate over UBI, sovereign wealth funds, and equity-sharing is not academic. It will determine whether the AI revolution creates a society of owners or a society of dependents.

**The Bottom Line:**

Anthropic has a concept of a plan for 25% unemployment. The plan is a triage manual, not a final answer. The $350 million funds research, not relief. And the question of whether the industry can be trusted to prepare for the disruption it is causing remains unanswered.

The “unprecedented” scenario may never arrive. But the fact that the architects of the AI revolution are preparing for it is a warning we should not ignore.

---

**#Anthropic #AI #JobDisplacement #UniversalBasicIncome #FutureOfWork #AIPolicy #DarioAmodei**

---

*Disclaimer: This article is for informational purposes only. It does not constitute financial or policy advice. The scenarios and proposals described are based on Anthropic’s published framework and are subject to change.*