Founder Power Splinters as AI Megadeals Propel Record $122B Quarter

When Doctolib disclosed in early April that employees and early investors had sold roughly €300 million in shares, the French health-tech company called it a milestone — proof of liquidity in a market starved for exits. The press release omitted one number: the discount. Secondary buyers acquired the stock at a valuation that implied a 38% markdown from Doctolib's last primary round, according to Bloomberg data reviewed by the buyers' advisors. A 38% haircut on a secondary is not a liquidity milestone. It is a price signal — and in a quarter when venture funding shattered every record, it was the price signal that mattered more than the headline numbers.

Global venture funding hit an estimated $285 billion to $300 billion in the first quarter of 2026, a figure that obliterates every prior quarterly record, according to CB Insights data published by Crowdfund Insider's Omar Faridi. Four U.S.-based megadeals — led by OpenAI's $122 billion raise at an $852 billion valuation — collectively accounted for roughly $188 billion of that total. Another way to read the data: a handful of companies absorbed two-thirds of all venture dollars deployed on the planet, while the remaining six thousand funded startups split what was left. The average check size hit records. The median startup saw its negotiating leverage evaporate.

The geographic data tells the same story in sharper relief. In the Seattle metro area, Q1 deal count dropped to 69 — the lowest number since mid-2020, according to the PitchBook-NVCA Venture Monitor, reported by GeekWire's Todd Bishop. Seattle ranked seventh in the country by total capital invested but tenth by deal count. The dollars that did land concentrated into a handful of names: Stoke Space, Temporal, XBOW, Overland AI. Pittsburgh saw a similar pattern: $1.7 billion in total funding, almost all of it driven by a single deal, as Technical.ly reported. The same dynamic appeared in Louisiana. Everywhere you looked, the denominator — the number of founders who could credibly run a competitive process — was shrinking.

AI companies captured 81% of all venture dollars in the quarter. This is not sector leadership. This is sector suffocation. Non-AI founders — in biotech, in digital health, in enterprise SaaS, in climate — spent Q1 confronting a market where lead investors would take the first meeting, ask about the AI roadmap, and then pass. The ones who raised did so on terms that looked closer to 2023 than to the euphoria the headline numbers implied. Participating preferred, once a relic of down-round cleansing cycles, crept back into term sheets for Series B and later-stage deals. Multiple liquidation preferences — the clause that lets investors take their money out more than once before common sees a penny — reappeared in restructuring rounds that weren't called restructuring rounds.

At the top of the market, the power dynamic was inverted entirely. OpenAI's $122 billion round, led by SoftBank with a syndicate that includes MGX, the Fundrise Innovation Fund, and a roster of sovereign wealth vehicles, was structured so favorably to the company that several participants waived pro-rata rights on future rounds to get into the cap table at all. When you are raising at an $852 billion valuation and your lead is a strategic investor with a compute-lease agenda, you do not give away board seats easily. You do not accept a participation cap. You dictate the docs. Sam Altman's team negotiated information rights that are narrower than what a Series B SaaS founder in Des Moines would have conceded to a Midwest seed fund in 2018.

The Seattle area ranked seventh in the U.S. by venture capital invested but 10th by deal count in the first quarter of 2026, as funding concentrated into fewer, larger rounds.— PitchBook-NVCA Venture Monitor, via GeekWire

Yet the OpenAI round is also where the split-screen effect reveals its sharpest tensions. Within weeks of the close, Reuters reported that some of OpenAI's own backers were questioning the $852 billion mark as the company shifted its commercial strategy toward enterprise customers to fend off Anthropic. The Financial Times had the internal growth numbers: revenue was tracking below the forecasts circulated during fundraising. This is the peculiar risk of the founder-dominant megadeal — when you raise at a price decoupled from operating metrics, the investors who supported you become the auditors of your next quarter's revenue miss. They cannot fire you, but they can refuse to re-up.

Anthropic, meanwhile, saw its own secondary-market valuation cross $1 trillion on some trading platforms, surpassing OpenAI by more than $100 billion, as the New York Post reported in late April. The two companies are now locked in what one GP at a multistage fund described to me as "a mutual-hostage situation with the public markets" — neither can go public at these numbers without inviting a correction that would shred the multiples of every AI-adjacent portfolio company, but neither can stay private indefinitely without testing LP patience on distributions.

Fund-of-funds analysts described a growing anxiety that has little to do with the AI thesis and everything to do with denominator math. When a $2 billion fund writes a $500 million check into a single company at an $852 billion valuation, the arithmetic of returning 3x to limited partners requires that company to eventually be worth more than the entire current market capitalization of Alphabet.

The career-risk calculus for GPs has shifted the negotiating table in ways that founders outside the AI bubble are only beginning to feel. When a venture capitalist is carrying a portfolio mark that requires a trillion-dollar outcome to justify, they become conservative on every other line item. Associates at three different Bay Area funds described their partnership's standard term sheet for non-AI Series A deals as including board-composition language that ensures the investor bloc can block a founder-CEO change, a sale, or a new fundraising round without unanimous consent. One firm introduced a clause requiring founder vesting to reset upon any down round — a provision that was virtually extinct during the 2024–2025 recovery window.

This is what the Q1 data does not capture: the quiet resurrection of terms that the bull market had supposedly banished. Founders raising between $10 million and $50 million are seeing preferred-stock provisions that give the lead investor a 2x liquidation preference on a sale below the post-money valuation, plus a participating feature that allows the investor to double-dip into the common pool. In plain English: if you sell the company for less than what everyone hoped, the investors get paid back twice before you see a dollar. This structure was common in 2016 and 2022; it all but disappeared in 2024. It is back.

Board-control dynamics are shifting in parallel. The standard founder-friendly term sheet of 2024 — two common board seats for the founders, one for the lead, one independent — is being replaced, deal by deal, by configurations that give the investor bloc effective veto power. The mechanism is often procedural rather than structural: quorum requirements that cannot be met without investor-director attendance, or a defined set of "protective provisions" that require a supermajority investor vote. The optics of a 2-1-1 board remain, but the functional control has moved.

The secondary market is where the tension resolves, and Doctolib's 38% discount is the template, not the exception. When employees and early investors sell in secondary transactions, the clearing price is often the only honest mark in the entire capital structure. Primary rounds are priced by negotiation between a motivated lead and a company that controls the data room. Secondary prices are set by a buyer who has no relationship with the founder and no incentive to flatter the last round's valuation. The spread between those two numbers — primary mark and secondary clearing price — is the most underreported number in venture capital, and in Q1 2026 it widened further than at any point since the 2022 correction.

For founders in the middle — the enterprise SaaS companies, the digital health platforms, the climate-tech hardware plays — the widening spread creates a structural disadvantage. When your company's secondary shares trade at a discount to the last round, recruiting becomes harder: candidate equity packages are valued at the primary price but discounted in the candidate's mental model. Retention becomes harder: employees who joined at the last strike price see underwater options. And fundraising becomes harder: the next lead investor will anchor on the secondary price, not your last post-money, and they will extract terms accordingly.

The lawyers who drafted the Q1 deal docs — I spoke with partners at two New York firms who handle venture work — describe a market where the term-sheet negotiation has become a proxy fight over who absorbs the valuation reset that everyone can see coming. Founders want the reset priced into the next round, where dilution is shared across all existing investors. Lead investors want the reset priced into the current round, via structured preferred instruments that protect the new money while leaving common and prior preferred to absorb the downside. The outcome depends almost entirely on whether the founder has an alternative offer. If she does — if there is genuine competition for the lead — she can push the reset downstream. If she doesn't, the term sheet arrives with a participation cap and a compounding dividend.

One piece of data from the CB Insights report that did not get enough attention: the number of venture funds actively deploying capital in the U.S. declined in Q1, even as total dollars hit records. The denominator of active GPs is shrinking. Fewer funds means fewer leads chasing each deal, which means less competition on terms, which means the balance of power tips further toward the surviving check-writers. The venture industry is consolidating just as the startup market is, and the consolidation has the same effect at the fund level that it has at the startup level: the large get leverage, the small get squeezed.

What should a founder take from a quarter that broke every aggregate record while leaving most individual founders worse off? The AI frontier is its own economy, and its gravitational pull distorts every metric that used to signal market health. Total funding is no longer a useful gauge of founder leverage. Deal count is. Term-sheet structure is. The spread between primary marks and secondary clearing prices is. Watch those three numbers in Q2. If deal count keeps falling while structure keeps tightening, the $300 billion headline will have been a mirage for everyone except the half-dozen companies whose cap tables it actually describes.