The RIF Detector

In February 2026, Jack Dorsey cut 4,000 people — nearly half of Block. He told investors AI was the reason and said the rest of the industry would follow. The stock jumped 25% the next day.

That number is real. But it raises two honest questions. First: does the market actually reward layoffs when a company blames AI, or does it reward something else entirely? Second: can we figure out which company is next?

So we built something we are calling the RIF Detector. "RIF" is just the corporate word for layoff — Reduction In Force. The goal was simple: look at every public company that did an AI-blamed layoff, check what the stock actually did, and compare it against companies that laid people off for other reasons.

How We Did It — In Plain English

If a company announces layoffs on a day the whole market is up 2%, you cannot give the layoff credit for the stock going up 2%. The market was already going up. So we stripped that out. For every layoff in our dataset, we measured how much the stock moved beyond what the market did that day. That is the only honest way to judge the news itself.

Then we split the layoffs into two groups:

• Group A — "AI is the reason": 16 companies from 2024–2026 that explicitly blamed AI or efficiency software for the cuts.
• Group B — everyone else: 12 large layoffs with normal reasons ("we overhired," "restructuring," cost cuts). Meta 2022, Spotify 2023, PayPal, Cisco, Zoom, and friends.

We pulled all the data from boring, reliable places: SEC filings, actual earnings call transcripts, Yahoo Finance. No SEO layoff trackers, no anonymous blogs. If a claim is in this post, it came from a document someone had to sign their name to.

Here is the part nobody is talking about. When you strip out the market noise, the "AI made us do it" layoffs actually underperformed the normal layoffs. Not by a little. By a lot.

On the day of the announcement, normal layoffs bumped the stock 5.6 times more than AI-blamed layoffs. And the gap gets wider over the following weeks. Look at what happens as time goes on:

This is the opposite of what most financial press has been writing. The narrative says AI layoffs are the new gold standard. The data says: not really. Not on their own, anyway.

So if the "AI" label alone does not make a stock pop, what does? Two things showed up clearly in the data.

1. A Credible Plan For The Money They're Saving

The #1 thing separating winners from losers was whether the company gave a real number for how much money they would save and what they would do with it.

Block's 25% pop was not about AI. It was because Dorsey said, clearly, that he expected the profit margin to jump by 6 percentage points and roughly $3.2 billion in new profit. Workday's bounce worked because they showed profit expanding from 25.9% to 29.6%. The market is not dumb — it wants a receipt, not a vibe.

2. A Technical CEO In Charge

The second thing: technical CEOs — the founders and engineers who came up through code — made cuts that the market actually believed.

Dorsey, Spiegel, Ellison, Cannon-Brookes — they can say "we're replacing a team with AI tooling" and investors buy it, because they actually write code. When a non-technical CEO says the same sentence, it sounds like they read it in a McKinsey deck.

The Full List (And What Happened)

Here is every public AI-blamed layoff in our dataset, sorted by how the stock reacted. You can see the pattern: plan + technical founder at the top, vague pain at the bottom.

✦ = founder or engineer CEO. Stock move = move after subtracting what the market did that day.

One More Thing That Does NOT Matter

We also tested whether how often a company says "AI" on their earnings calls predicts anything. It does not. Zero predictive power. Saying AI in public and actually cutting headcount are unrelated activities.

Before we get to predictions, we need to admit something. Our first draft of this research had a big error, and the story behind it is worth telling on its own.

We used Wolters Kluwer — the legal and tax software giant in Europe — as a headline example of a company ripe for AI-driven cuts. The number we put in the first draft was that they had about 6,000 engineers and were building a 100-person AI platform team to start replacing them. That ratio made them look like a layoff waiting to happen.

Here is where the story gets interesting. The 6,000 number did not come from us guessing. It came from a Wolters Kluwer recruiter, on a call. Someone who actually works at the company, sourcing candidates for engineering roles, told us that was their engineering headcount.

We took it at face value. That was the mistake.

Rule we should have followed the first time: even if the person on the other end of the phone works at the company, cross-check the number against a primary document before publishing it.

When we went back and pulled LinkedIn employee data, WK's own annual report, and third-party engineering-headcount trackers like Unify, the real number was closer to 1,210 engineers. The recruiter was probably counting product managers, designers, data people, and some IT operations staff as "engineers." That is a normal thing for a recruiter to do — their definition of "the engineering org" is much looser than a headcount analyst's.

Once we corrected the denominator, the picture flipped. A company with 1,200 engineers and a 100-person AI team is not scrambling — they are investing at a reasonable pace. Combined with a 32% operating margin (among the healthiest in the sector), Wolters Kluwer dropped from "almost certain to cut" down to "might, in a year or two, if things get worse."

We are leaving this whole mistake in the post on purpose. Financial research that privately corrects itself and publicly projects confidence is exactly how bad Wall Street analysts work. If we got something wrong, you should see it.

OK. With that out of the way — the whole point of this research is to find companies before they announce layoffs, so the pop happens while we already own the stock, not after.

The pattern that worked for Block, Snap, Oracle, and Workday is this: a company under real pressure (activist investor, weak margins, peers already cutting), led by a credible CEO, that can announce a layoff with a specific savings number attached. When that combination shows up, the stock jumps 5% to 25% on the news.

Here are the public companies where all three ingredients already exist. The layoff has not been announced yet. The pressure is already on the table.

Cap = market capitalization. Likely Cut = estimated headcount reduction the company would announce. If It Hits = expected one-day stock move IF the layoff is announced with a credible savings plan. If the announcement comes without a plan, expect the opposite direction.

The Reasoning, Company By Company

OK so you see a list of tickers. Now what? There are three ways people can act on a setup like this, with very different risk profiles.

Option 1 — Just Buy The Stock

The most boring and the most forgiving. You buy shares at or near the entry range. If the layoff is announced with a credible plan, you get the +15% to +40% pop. If nothing happens for a year, you sit with it. If the thesis breaks (activist leaves, margins improve without a cut), you sell flat or slightly down.

Good for: people who don't want to babysit. Risk is limited to your purchase price.

Option 2 — Buy A Basket

Buy small positions (0.5% to 1% each) across 8 to 12 of the names. Some will pop, some will sit, some will go down. The math actually works here:

This assumes unhit names average roughly flat. Real-world basket returns will depend on timing, position sizing, and how correlated the cuts are. But as a framework, a diversified basket of 8–12 AI-pressure candidates has a reasonable shot at double-digit returns over 12 months, with far less single-name risk.

Option 3 — Options (Very Different Game)

Buying call options on these names is tempting but tricky. Two things work against you:

• Implied volatility is already elevated. When activists are circling and short interest is high, options are expensive. You're paying a premium for the pop that might not happen in your window.
• IV crush after the announcement. The moment the layoff gets announced, the uncertainty that inflated the option price collapses. Even if the stock pops 20%, the option can gain less than you expect because the implied volatility drops at the same time.

If you go this route, longer-dated (6+ month) out-of-the-money calls are the typical way to play it. Shorter-dated options are lottery tickets — you need the exact week right.

Exit Rules (For All Three)

• Layoff announced WITH a savings plan: Take profit on 50% of the position within 48 hours of the pop. Let the rest run for 2 weeks and re-evaluate.
• Layoff announced WITHOUT a plan: Sell immediately. The stock will probably fade over the next two weeks (Pinterest fell 9.6% after their AI-blamed cut with no margin guide).
• Activist sells, company gets acquired, or management reshuffles: Usually exit — the original thesis is broken.
• A year goes by with nothing: Re-evaluate the thesis. If the pressure is still there, hold. If the activist left or margins improved another way, close the position and move on.

Here is what this research cannot do, in plain English.

• The sample is small. 31 companies is not a lot. We can see strong patterns, but we cannot tell you the exact odds with any kind of confidence.
• It cannot predict the exact day. We can tell you a company looks like a layoff candidate. We cannot tell you it will announce in March versus July. Use this to build a watchlist, not to time options.
• The world moves fast. AI itself is getting better every few weeks. A pattern from 2024 might be too conservative by late 2026. We're looking in a rearview mirror at a road that's changing shape.
• Correlation vs. cause. We can see that technical CEOs do AI-branded cuts more often. We cannot prove their technical background is why. It might just be that the same companies that hire technical CEOs also happen to be the ones that overhired the most during the 2021 boom.
• Dead companies don't file. Some firms cut people, blamed AI, and still went bankrupt. They're not in our data because they stopped existing. That makes the results look slightly rosier than reality.

Block's 25% pop was real. But Dorsey's quote — "within the next year, the majority of companies will reach the same conclusion" — was not entirely honest about why his own cut worked.

His cut worked because he is a technical founder who handed investors a receipt for the savings. That combination is rare. Most companies saying "AI efficiency" on their next earnings call will have neither. The market will figure that out, post by post, layoff by layoff.

Our full watchlist for the next few quarters: APPN, ASAN, DOMO, WK, BILL, TASK, CNXC, FISV, GCI, EPAM, SPT, BIGC. These are companies where the pressure is real, the leadership structure is in place, and the AI story is likely to show up in an 8-K filing, not just a tweet.

We'll update this post as announcements happen. If we're wrong, we'll say so — the same way we said it about Wolters Kluwer.

Reminder: none of this is financial advice. The tables, entry ranges, and potential returns above are speculative estimates based on historical patterns in a small sample. Past patterns do not guarantee future results. Consult a licensed financial advisor before making investment decisions.