Founders bet everything on being the one to break through. They raise, build, hire, and push through years of uncertainty on the belief that their company will be the one to win.
In the first half of 2026, we supported more dissolutions than in the same period last year, and saw the fewest AI company closures in our company history.
That’s different from what some headlines would predict. With AI dominating the market, it would be natural to assume AI shutdowns would correlate. But something else is driving the volume, and our data points to the SaaSpocalypse: the wave of pre-AI software companies now reaching the end of their runway.
At SimpleClosure, we support companies at the end of their journey. This year, we looked at the data from every shutdown we’ve handled between January and June 2026 to understand what’s happening in the ecosystem.
Let’s dig in.
The AI reckoning has not reached the shutdown data
AI took 86 cents of every US venture dollar in H1 2026. It accounted for 14.4% of the companies we closed, even as the total number of shutdowns we supported has drastically increased.
AI companies accounted for $355.9 billion of the $412.7 billion invested in US startups in the first half of 2026, yet they were just 14.4% of the companies that closed through SimpleClosure in the same window.
That share is falling against a rising total. AI represented 17.7% of our closures two years ago and 15.9% in our December 2025 report. Meanwhile, the number of companies we’ve helped close has grown substantially over that same stretch. AI’s share is shrinking, but not because fewer AI companies are shutting down. Other sectors are simply seeing more closures.
The median incorporation year for AI companies was 2024, versus 2022 for all other companies, making the median age at closure appear shorter for AI companies: 2.30 years versus 3.78.
We hypothesize that the 2023-2025 AI cohort is still in its early days, and we’re only beginning to see how it plays out.
AI companies stopped with a median of $30,000 in the bank, more than double the $13,000 median for the non-AI cohort. 91% still had a cash balance, which signals that they’re stopping before the runway ends.
One likely reason: AI founders have somewhere else to go. According to PitchBook, acquihires in AI and machine learning nearly doubled last year, from 181 deals in 2024 to 322 in 2025. In these deals, the team is typically the big draw, and the business itself is close to incidental. For a founder still holding cash, taking an acquihire offer can beat spending down the last reserves.
Of companies that raised a seed round of $1 million or more in 2023, only 24% made it past seed. Through 2020, the historical rate was 55% or higher.
Is there any value left to recover?
Often yes, and it is rarely cash. By the time a company decides to close, the bank balance is the least interesting line on the balance sheet. What still holds value is what the team built: the codebase, workspace and internal tooling, models and data, domains, and hardware.
Historically, the value of these assets simply evaporated. Servers were switched off, repositories archived, laptops left in a closet.
We built Asset Hub to change that. Asset Hub gives a dissolving company a structured way to sell what it still owns during the wind-down, rather than writing it off, and routes the proceeds back through the closing process. The market for those assets is real, and AI labs are among the most active buyers.
Timing is the constraint founders underestimate. Assets are worth the most while the company is still operating and the team can answer questions, transfer credentials, and document what they built. Once the entity is closed, most of that value is unrecoverable. Selling is a decision that belongs early in the wind-down.
For a company returning nothing to its investors, its assets are the last remaining source of value, and they are worth the most before the lights go out.
The SaaSpocalypse is real, and the casualties are not AI companies
B2B SaaS companies account for 27.3% of the shutdowns we supported, nearly twice the share of AI companies. AI has threatened a “SaaSpocalypse” for three years, and our dissolution data suggests it may already be taking shape. The SaaS companies shutting down were largely built before AI arrived. The AI-first companies aren't the ones closing.
These are older companies with a different profile in every aspect. The median B2B SaaS company we closed was incorporated in 2022 and closed at 3.78 years. More than half of our SaaS closures predate 2023 compared to just one in five AI closures. This generation of software companies raised and scaled before the platform shift and has struggled to fundraise.
In our data, AI-first and vertical industry categories do not overlap. Our industry categories are mutually exclusive and each company is counted once. A company building AI-native software is classified as AI, not as SaaS, so the SaaS and AI figures share no companies. AI, technology and software companies in aggregate account for 59% of the cohort: B2B SaaS (27.3%), AI (14.4%), Software and Technology (5.5%), Fintech (5.1%), Crypto and Web3 (2.9%), HealthTech (2.7%) and BioTech (2.0%).
Founders choosing to shut down is just as common as those running out of money
Among companies that told us why they closed, insufficient capital or cash flow problems accounted for 28.1%. A personal decision or change in priorities accounted for 22.2%, and founders' conflict or burnout for another 5.4%. Combined, founder choice reaches 27.6%.
The difference between shutting down compliantly and not is meaningful in time, cost, reputation, and burden. A company that runs out of cash enters dissolution with no resources to execute it properly, which is how founders end up with unfiled franchise tax returns, uncancelled EINs, and open foreign registrations that surface years later. A company whose founder chooses to stop while solvent has the ability to close in an orderly way: to notify creditors, settle obligations, and return remaining capital to investors according to the terms those investors actually signed.
This matters because the difference between closing in order and closing halfway is measured in years of personal exposure. Delaware still expects a franchise tax return. The IRS still expects the final filings that apply to the entity. States where the company registered still carry it as active until it withdraws. Inaction accrues obligation quietly, and it usually surfaces at the least convenient moment: a background check, a board reference, or diligence on the next company.
Inside the 27.6% of founders electing a shutdown, a change in priorities accounts for 22.2%, four times what conflict or burnout accounts for on its own. Most of what we're calling ”founder choice” is a founder deciding the company has run its course, and closing it while there's still something left to close in an orderly way.
The calculation founders are running is one our founder and CEO, Dori Yona, described, "If they are not AI-first, they are having a really hard time fundraising right now. It's sometimes easier for a SaaS company to just shut down and restart."
Closing is increasingly a decision solvent founders make on their own terms. A founder who intends to build again has a direct interest in closing the last one cleanly.
Our YoY Shutdown Analysis
Companies aren’t necessarily dying faster than they were a year ago.
What’s happening instead is slower and less visible. These companies didn’t run out of runway overnight. They went two years without another raise, with no next round to extend it.
The numbers did not move significantly year over year. Median time from last raise to closure was 2.28 years in both H1 2025 and H1 2026. Series A companies fell slightly as a share of closures, from 14.0% to 11.2%.
The money itself has moved, and mostly to one place. Megadeals of $100 million or more captured 87.5% of the $412.7 billion invested in US startups in H1 2026, leaving 12.5% to cover every seed, Series A, and Series B deal combined. Anthropic alone raised $65 billion in a single round that quarter, more than the total early-stage capital raised across the entire country.
The market has become more selective in a way that compounds the problem. A company that raised its last round in 2022 or 2023, before AI reshaped what investors expect, is now being measured against a different bar: faster growth, leaner headcount, product built around AI from day one. The company itself may not have changed much. What was enough to raise in 2022 or 2023 may no longer be enough today.
What to look out for
The AI cohort now inside its first runway will reach its outcomes over the next 24 to 36 months. Watching whether the share of AI shutdowns starts to climb, or keep falling, is one of the clearest signals we’ll have.
