The $100 Billion Carcass

Thoma Bravo has created more opportunity for AI startups than Y Combinator. They just didn’t mean to.

Right now in Manhattan, a software CEO is watching his company die in a conference room. He knows exactly what’s killing it. His VP of Product showed him Sierra’s numbers six months ago — an AI agent that resolves customer issues for less than a dollar. A price point his cost structure will never touch. He could rebuild the product from scratch. The lenders who funded the buyout have him in a cage. He could acquire the replacement. His balance sheet can’t support it. So instead he’s spending his last dollars on consultants and restructuring advisors who are negotiating the terms of his surrender — while a 26-year-old founder in Hayes Valley is on the phone with his biggest customer, offering to do what his company no longer can.

That CEO runs one of nearly 2,000 software companies that private equity loaded with debt over the last decade. The PE firms — Thoma Bravo, Vista Equity, Francisco Partners, Clearlake — took healthy companies with loyal enterprise customers, strapped on more debt than the businesses could carry, stripped R&D to make the interest payments, extracted over $20 billion in fees, and walked away before the bill came due. The capstone: Vista and Elliott bought Citrix for $16.5 billion — the largest software buyout in history. Even the crown jewel is underwater.

Then the market turned. Interest rates rose and stayed up. Software valuations cratered over 70%. Debt went from 30% of company value to 70%. At that ratio, the lenders own the company. They just haven’t taken the keys yet. And then AI broke the only assumption holding the rest together.

The bill is now due. And the companies cannot pay it.

More than $120 billion in software debt matures in the next four years. The peak hits in 2028 — eighteen months from now.

The lenders who funded these buyouts are already pricing in losses — over 60% of the debt is trading at a deep discount. The companies underneath cannot refinance, cannot invest in product, and cannot compete with anyone who doesn’t carry their debt load. The equity is zero.

And they cannot adapt. All of this debt depends on one assumption: seat-based software subscriptions are permanent. More employees at the customer means more seats means more revenue means the debt gets paid. AI agents break that equation permanently. When an AI agent makes a customer 3x more efficient, that customer doesn’t need 300 seats anymore. They need 100. The revenue the debt depends on evaporates. And the debt agreements won’t let them change their pricing to fight back.

The balance sheet is a legal prohibition against adapting to AI.

The seven most valuable technology companies on earth spend a combined $700 billion a year on AI capex with almost no net debt. They could borrow three times their earnings tomorrow. They choose not to. Because they understand something PE sponsors never figured out: you can’t fund a rebuild when every dollar goes to debt payments.

The bodies are already on the ground. Count them.

Avaya was the backbone of enterprise communications — the phone system in every Fortune 500 lobby, the contact center platform behind every customer service call. PE loaded it with debt. It filed for bankruptcy. Restructured. Filed again. The second time around, the people who lent them money got back 24 cents for every dollar. While Avaya was drowning, Bret Taylor — former Salesforce co-CEO — launched Sierra to handle customer conversations for less than a dollar each. $100 million ARR in under two years. Zero debt. A pricing structure that Avaya, buried under its own balance sheet, is physically incapable of matching. That is what carnage looks like when a founder lands on a carcass first.

Anthology — the successor to Blackboard — ran the learning management system inside nearly every major university in America. Two decades of dominance. When its debt started cracking, Wall Street told the lenders they’d get nearly all their money back. Eighteen months later: Chapter 11. Lenders got back 0.2 cents — a fifth of a penny for every dollar they thought was safe. The EdTech startups that replaced it were already funded before the bankruptcy was even filed.

McAfee — the name that was on every PC in America — buried under PE debt, bleeding market share to AI-native security platforms. GoTo, the company that powered remote work before Zoom existed, restructured and still sinking. RSA, the encryption standard for two decades, gutted by its sponsors. Sandvine — the investors who funded that buyout got back pocket change.

These companies are not failing because their products stopped working. They are failing because their debt won’t let them compete.

The restructuring advisors bill $2,000 an hour to rearrange deck chairs. The founders replacing their clients’ products charge $0.99 per resolution.

And the hedge funds showing up to “save” what’s left are making it worse. Elliott, Oaktree, Sixth Street — they buy the debt at a discount and need massive returns on a short timeline to make the trade work. That math only works through cost-cutting. When they take control, R&D gets gutted further. Customer success disappears. The product rots. The customers get angrier. And every quarter, the AI-native competitor’s win rate ticks up.

The Midtown vultures aren’t competing with the founders. They’re tenderizing the meat.

The first founders found the carcass without a map. The next wave won’t need one — the bodies are impossible to miss.

Bret Taylor saw Avaya drowning under its own debt and built Sierra to take every customer Avaya could no longer serve — $100M ARR in under two years. NinjaOne watched PE gut the R&D budgets at Ivanti and McAfee and went straight at their $7 billion in trapped IT customers — $500M+ ARR and accelerating. Wiz didn’t wait for the security incumbents to die. It just built the product they couldn’t, while their PE sponsors were busy extracting fees. $32 billion valuation. Rippling at $17 billion and Ramp at $32 billion are dismantling the mid-market ERP stack that PE froze in place. Abridge raised $300 million in a Series E to automate clinical documentation that legacy healthcare software can no longer improve.

None of these founders studied the debt markets. They didn’t need to. They built products so good that the incumbents were too paralyzed to respond. But here’s what should keep you up at night: the companies they’re displacing aren’t just losing. They are contractually prohibited from fighting back. Their debt agreements literally prevent them from investing in a rebuild. Every quarter the product degrades further, the customers get more frustrated, and the door opens wider.

Every day you spend building against a company with a clean balance sheet and ten thousand competitors is a day someone else is eating the revenue that’s sitting here undefended.

The feast is accelerating. And there is still $100 billion of carcass left.

The Target Map

The aerial view. Six feeding grounds. Over $30 billion in software revenue with no one left to defend it. The incumbents are trapped by debt, starved of R&D, and bleeding customers. The vultures already circling each one are named below. If yours isn’t on the list yet, the window is still open.

If you recognize your market in the map above, the window is open.
If you recognize your portfolio, it may already be closing.

Build Accordingly

Every platform shift in the history of technology has been won by small teams with clean balance sheets, moving fast into markets that incumbents could see but could not reach. The PC revolution was not won by IBM. The internet was not won by AOL. Mobile was not won by Nokia. The AI transition will be won by founders — backed by VCs who understand that the opportunity is not speculative. It is measurable. It is named. And it is on a ticking clock.

PE got paid. Everyone else gets the bill. The lenders. The LPs. The pension funds who were told this was “stable income” — it was a seat-based software contract that an AI agent can replace for pennies. The employees whose R&D budgets were redirected to interest payments. The equity is zero.

For decades, venture capitalists bristled at the “vulture capitalist” label. They should embrace it. The Midtown vultures — Elliott, Oaktree, Sixth Street — are circling the debt at 40 cents, extracting fees, and accelerating the collapse. The real vulture capitalists are the ones circling the customer bases those restructurings leave exposed.

When a Midtown vulture takes control, value gets extracted. When a Sand Hill vulture takes control, value gets created. Same instinct. Opposite outcome. The founders are the beak.

Y Combinator funds 500 startups a year. Thoma Bravo funded their market. The wall hits in 18 months. Build accordingly.