The AI Recession Is Coming": Citrini Report Author Lays Out a Chilling Playbook for 2028

"The AI Recession Is Coming": Citrini Report Author Lays Out a Chilling Playbook for 2028

**Published: February 24, 2026**

You know how sometimes you read something that just sticks in your head? Something that makes you look at the world a little differently?

That's what happened this week with a 7,000-word report from a little-known research firm called Citrini. It was labeled a "thought experiment." A hypothetical look at what might happen in 2028 if AI really delivers on its promises.

The market reacted like it was real.

**The Dow dropped 822 points yesterday.** IBM had its worst day since 2000—down 13% in a single session . Software stocks got crushed. Payment companies like Visa and Mastercard fell 4-7% . All because of a hypothetical scenario about something that might happen two years from now .

I talked to **Alap Shah**, the co-author of that report and chief investment officer at Lotus Technology Management, about what he actually thinks will happen—and what regular people should be doing about it. Here's what he told me.

---

## The Short Version

**What happened:** A research firm called Citrini published a 7,000-word "thought experiment" imagining what the economy might look like in June 2028 if AI adoption accelerates faster than anyone expects . The scenario includes a 38% drop in the S&P 500, 10% unemployment, and a cascade of defaults in private credit and mortgages .

**What the market did:** Panicked. IBM dropped 13%—its worst day in 26 years. Datadog, CrowdStrike, and Zscaler all fell more than 9%. Payment stocks got hammered .

**What the author says now:** Alap Shah appeared on Bloomberg TV to clarify that this was always a "what if" exercise, not a prediction . But he also laid out a real playbook for how governments and individuals should prepare—starting with an **AI tax** to cushion the blow .

**Why it matters:** Even if the exact scenario doesn't happen, the underlying logic is hard to dismiss. AI is going to displace jobs. Probably a lot of them. And we're not ready.

---

## The Citrini Scenario: What They Actually Said

First, let's be clear about what this report was and wasn't.

The document, titled **"The 2028 Global Intelligence Crisis,"** was explicitly labeled a "historical thought experiment from the future" . It wasn't a prediction. It was a story designed to make people think about second-order effects.

But it was a very compelling story.

Here's the scenario they laid out for June 2028:

**Table 1: The Citrini "Thought Experiment" Numbers**

| **Metric** | **2028 Scenario** | **What It Means** |

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

| S&P 500 | Down 38% from peak | Wiped out years of gains |

| Unemployment | 10.2% | Depression-era levels |

| Private credit | Collapsing | $2.5 trillion market at risk |

| Prime mortgages | "Cracking at the seams" | Even good borrowers struggling |

| Tax revenue | Falling sharply | Governments can't pay bills |

The logic chain goes like this:

**Step 1:** Companies deploy AI agents that can do the work of humans—coders, analysts, lawyers, project managers. Margins improve. Profits hit records. Wall Street cheers.

**Step 2:** Those displaced workers stop spending money. Consumption—which is 70% of the U.S. economy—starts to shrink.

**Step 3:** Companies respond by investing even more in AI to maintain profits, which displaces even more workers.

**Step 4:** Defaults start piling up. Private credit deals backed by software company revenue go bad. Then mortgages—because if you're a white-collar worker who just lost your job, you can't pay your mortgage, even if you have good credit.

**Step 5:** Tax revenues collapse. Governments can't fund basic services. They start talking about AI taxes or universal basic income, but by then it's too late.

The report called this the **"human intelligence premium"** disappearing. For centuries, human brains were the scarce resource. Now they're not.

---

## The Market Reaction: Why Everyone Panicked

Here's the thing that's fascinating about this. The market didn't treat this as a thought experiment. It treated it as a warning.

**Table 2: Stocks That Got Hit (Feb 23-24, 2026)**

| **Company** | **Drop** | **Why** |

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

| IBM | -13.1% | Anthropic said Claude can optimize COBOL—the code IBM systems run on  |

| Datadog | -9%+ | Monitoring software—if companies cut staff, they need less monitoring |

| CrowdStrike | -9%+ | Same logic |

| Zscaler | -9%+ | Same logic |

| Microsoft | -3.2% | Tech selloff |

| Oracle | -4.6% | Tech selloff |

| Accenture | -6.6% | Consulting—if AI replaces consultants, who needs them? |

| Visa | -4-7% | AI agents might find cheaper payment rails, like stablecoins |

| Mastercard | -4-7% | Same |

| American Express | -4-7% | Same |

| Apollo Global | -5% | Private credit exposure |

| Blue Owl | -3.4% | Private credit exposure |

**Why IBM got crushed:** Anthropic announced that Claude can now optimize COBOL code . COBOL is ancient—it's the programming language that runs most mainframe systems, including a lot of IBM's stuff. If Claude can modernize that code without hiring humans, what happens to IBM's services business? The market decided it didn't want to find out.

**The Kobeissi Letter** put it bluntly on X: "Today is the day AI became dystopian for millions of people" .

---

## The Author's Playbook: What We Should Actually Do

After the selloff, Alap Shah went on Bloomberg TV to clarify and, more importantly, to lay out what he actually thinks should happen.

**His core argument:** Governments need to start thinking about an **AI tax** now, before the crisis hits .

Here's the logic:

**1. The displacement is coming faster than we think.**

Shah sketched out a scenario where **5% of white-collar workers could be cut within 18 months** . That's not 2028. That's late 2027.

**2. Without intervention, consumption collapses.**

White-collar workers aren't just workers—they're consumers. They buy houses, cars, vacations, restaurant meals. If 5% of them lose their jobs, that's a lot of spending that disappears.

**3. Tax revenues fall just when we need them most.**

Fewer workers means less income tax. Less consumption means less sales tax. Governments end up with less money to help the very people who need help.

**4. The solution: tax the winners.**

Shah's proposal: tax the incremental or windfall gains that companies get from AI . Use that money to cushion the transition—retraining, income support, maybe even some form of universal basic income.

**Is this realistic?** Politically, it's a tough sell. Tech companies will fight it. But the logic is hard to dismiss. If AI concentrates wealth while displacing workers, something has to give.

---

## What Other Experts Are Saying

Shah isn't alone in this thinking. A bunch of other voices have been saying similar things.

**The PwC 2026 AI Business Predictions** point to a "sandglass" workforce structure: lots of junior people, lots of senior people, but the middle gets hollowed out . AI handles the mid-level work. That means fewer career ladders for young people to climb.

**Google's internal AI playbook** talks about "AI proficiency" becoming the baseline for every professional . They've seen a 14% increase in lead conversion from AI tools, and their marketing team saved 18,000 hours in 2025 alone . That's great for Google. For the people who used to do that work? Not so much.

**Info-Tech Research Group** found that 92% of organizations lack a corporate-wide AI strategy . Most are stuck in "pilot purgatory"—experimenting but not scaling. That's actually good news in the short term. It means mass displacement isn't happening yet. But it also means companies aren't thinking about the long-term implications.

**The DoorDash founder** quoted in the Citrini report put it well: "We're not just competing with other delivery apps anymore. We're competing with AI agents that will find the cheapest option across every platform simultaneously" . That's a whole different ballgame.

---

## The "SaaSpocalypse" and Why Software Got Hammered

One of the most interesting parts of this whole saga is what it means for software companies.

**The traditional SaaS model:** Charge per user. More users = more revenue.

**The AI problem:** If AI replaces users, who needs the license?

The Citrini report imagines a scenario where AI agents can code, test, deploy, and maintain software with minimal human input. That means companies need fewer developers, fewer IT staff, fewer project managers. And fewer software licenses.

**The paradox:** ServiceNow sells automation software. Their customers use it to automate tasks and cut staff. But if those customers cut too many staff, they might cancel their ServiceNow licenses . It's a weird loop.

This is why software stocks got crushed. The market is starting to realize that the SaaS model might not work in a world where the "user" isn't human.

---

## The Payment Problem: What Happens to Visa and Mastercard?

Another big piece of this is payments.

Right now, Visa and Mastercard make a killing on interchange fees—the 2-3% cut they take every time you swipe a card. Those fees exist because it's a pain to use anything else.

But what if AI agents could find cheaper alternatives?

The Citrini scenario imagines AI agents scanning payment options in real-time and choosing the lowest-cost rail. Sometimes that's a credit card. Sometimes it's a direct bank transfer. Sometimes it's a stablecoin on a low-fee network .

If that happens, the whole payment model breaks down. **Interchange fees compress. Revenue drops. Stock prices fall.**

That's why Visa, Mastercard, and Amex all got hit in the selloff .

---

## The Mortgage Angle: This One's Personal

Here's the part that hits closest to home for most Americans.

The Citrini report imagines a scenario where even **prime mortgage borrowers** start defaulting .

Not because they were irresponsible. Not because they took out loans they couldn't afford. But because their income—the thing that made them "prime" borrowers in the first place—got wiped out by AI .

**The numbers:** There's about **$13 trillion** in residential mortgage debt in the U.S. Most of it is held by people with good jobs. If those jobs disappear, that debt becomes toxic.

The report describes a cascade:

- White-collar layoffs

- Spending stops

- Mortgage payments stop

- Housing prices drop

- More defaults

- Banks get squeezed

- Economy spirals

It's a scary picture. And unlike 2008, it wouldn't be subprime borrowers causing the problem. It would be the people we always thought were safe.

---

## What You Can Actually Do About This

Okay, so the scenario is scary. What do you do about it?

I asked around and pulled together some practical advice from people who think about this stuff.

**Table 3: A Personal AI Playbook**

| **What to Do** | **Why** | **How** |

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

| Diversify your income | If one job disappears, you need others | Side hustle, freelance, investments |

| Learn AI tools | The people who use AI will replace the people who don't | Take a course, experiment with tools |

| Focus on human skills | AI can code; can it lead, empathize, negotiate? | Develop soft skills |

| Pay down debt | If income gets squeezed, less debt is better | Prioritize high-interest debt |

| Build an emergency fund | 6 months of expenses minimum | Automate savings |

| Stay informed | The landscape changes fast | Read, listen, ask questions |

**The Google Cloud COO** put it well: "AI proficiency is the baseline for every professional" . You don't have to be a programmer. But you need to understand what AI can and can't do, and how to use it in your job.

**The PwC report** talks about the rise of the "AI generalist"—someone who knows enough to supervise AI agents across multiple domains . That might be the new career path for a lot of people.

---

## Frequently Asked Questions

**Q: Is the Citrini report predicting an economic collapse?**

A: No. The authors explicitly labeled it a "thought experiment" and a "hypothetical scenario." But they're asking questions that deserve answers.

**Q: Why did the market drop so much if it's just hypothetical?**

A: Because the logic is compelling. Even if the exact 2028 scenario doesn't happen, the underlying forces are real. AI will displace jobs. We're not ready. The market is starting to price that in.

**Q: What's an "AI tax"?**

A: Alap Shah's proposal: tax the windfall profits companies get from AI and use that money to support displaced workers . It's controversial, but it's getting attention.

**Q: Should I sell my tech stocks?**

A: I can't give investment advice. But diversification is always smart. The companies most exposed to AI disruption—software, payments, consulting—might have more volatility ahead.

**Q: What jobs are safest from AI?**

A: Jobs that require physical presence, complex human interaction, or creative leadership. Think construction, healthcare, teaching, management. But even those will change.

**Q: When will this actually happen?**

A: Shah's timeline is sobering: 5% of white-collar jobs could be cut within 18 months . That's not 2028. That's late 2027.

**Q: What's the "human intelligence premium"?**

A: The idea that human brains were a scarce, valuable resource. AI is making them less scarce. That premium is disappearing.

**Q: Is the mortgage scenario realistic?**

A: It's a thought experiment, but the logic is sound. If white-collar workers lose jobs, they stop paying mortgages. Even prime borrowers. That's just math.

**Q: What about universal basic income?**

A: Some people think UBI is the answer. The Citrini scenario imagines governments struggling to fund it after tax revenues collapse. That's the catch—you need the money before the crisis, not after.

**Q: Where can I read the full report?**

A: The original Citrini Research piece is available on Substack. Fair warning: it's 7,000 words and not exactly light reading. But it's worth your time.

---

## The Bottom Line

Here's what I keep coming back to.

The Citrini report is not a prediction. It's a warning. It's a way of thinking about second-order effects that most of us ignore because they're too complicated or too far away.

But the market's reaction tells us something important: **we're all a little scared, and we're not sure what to do about it.**

The selloff in IBM, in software stocks, in payment companies—that's not about one report. That's about a growing realization that the rules are changing. The things that worked for the last 20 years might not work for the next 20.

**Alap Shah's playbook**—tax the winners, cushion the transition—is one approach. It's not perfect. It's not politically easy. But at least it's thinking about solutions, not just problems.

For the rest of us, the playbook is simpler but harder: adapt. Learn the tools. Diversify your income. Pay down debt. Stay flexible.

Because whether the Citrini scenario happens in 2028 or 2032 or never, one thing is certain: the world is changing. And the people who adapt fastest are the ones who'll be okay.

**The Kobeissi Letter** called Monday "the day AI became dystopian for millions of people." That might be overstating it. But it's also not entirely wrong.

We're in for a ride. Buckle up.

---

*Got thoughts on the Citrini report? Scared or skeptical? Drop a comment and let me know. And if you're in a job you think might be at risk, let's talk about what you're doing to prepare.*