NY's data center freeze | PitchBook News Summary

1. Key Themes

AI Infrastructure Is Hitting a Political Wall

New York's one-year moratorium on new data centers exceeding 50 megawatts marks a watershed moment for AI infrastructure investment. The constraint is no longer technical — it's governmental.

"Transformers and interconnection queues: Capital can eventually engineer around [that]. Permits and a governor's signature, it can't." — Harrison Rolfes, PitchBook Senior Analyst

"The headline is a ban. The actual event is that 'time-to-power' just became a valuable commodity in the AI stack. A state legislature, not a chip shortage, is now setting the price." — Harrison Rolfes

State Regulation Is Creating a Race-to-the-Bottom Among Geographies — With No Clear Winner

Even traditionally business-friendly states are responding to public pressure, leaving data center investors without a reliable alternative jurisdiction.

"It's not at all clear which states are going to be data center central. And when you're starting to see New York and Texas on the same side of an issue, it's an extraordinarily weird situation." — Ted Brandt, CEO of Marathon Capital

"A May Gallup poll found that 71% of Americans oppose a data center being built near them, outpacing the 53% of Americans who oppose a new nuclear power plant in their backyard."

The Independent Sponsor Model Is Outperforming Traditional Buyouts

Deal-by-deal investing is shedding its stigma and generating market-beating returns, aided by regulatory tailwinds and a PE talent exodus.

"The term 'fundless sponsor' has long had a hint of failure about it, the sense that a sponsor is without a fund because they were unable to raise one. Today, not only are these sponsors outperforming the market — at least according to one recent study — but favorable regulation, less competition and a widening pool of interested backers are making going 'fundless' look like the better option for many."

"Those deals generated a median equity IRR of 23.8%, compared with 18.5% for comparable investments made by US buyouts." — Institute for Private Capital, UNC Kenan-Flagler Business School

AI Infrastructure Investment Is Shifting Toward Fewer, Larger, More Targeted Bets

Capital is concentrating in expansion-stage deals focused on solving concrete bottlenecks — energy, semiconductors, logistics — rather than broad platform plays.

"Expansion-stage deal value rose in 2025, driven more by larger transactions than by higher volume. Median deal size doubled in energy and commercial transportation, rising from $14.8 million to $30 million and from $15 million to $30 million, respectively. That suggests investors are making fewer, more targeted investments to ease infrastructure bottlenecks, strengthen supply chains, and expand industrial capacity for broader digital transformation." — Deloitte Road to Next

Agtech's Resilience Has Run Out

After a period of outperforming broader venture during the downturn, the agrifood sector has capitulated — across all three verticals.

"A year after the agtech sector defied the downturn, VC dealmaking fell to multiyear lows across all three agrifood verticals: agtech, foodtech and consumer packaged foods."

2. Contrarian Perspectives

"Fundless" Is Now the Smarter Structural Choice for Many PE Operators

The conventional wisdom is that independent sponsors are second-tier players who couldn't raise a fund. The data says otherwise. With a 23.8% median equity IRR vs. 18.5% for traditional buyouts, and Washington now lifting SBIC leverage caps to $250M, the independent sponsor model is becoming structurally advantaged — not a consolation prize.

"At the same time, Washington has boosted the flow of investor capital into the independent-sponsor market almost unnoticed. The Investing in All of America Act, signed May 19, lifted the leverage cap on small business investment companies... to $250 million from $175 million, 'a lot of which will go into independent sponsor deals.'" — Brett Palmer, President of the SBIA

"Both talent and capital are being nudged toward the deal-by-deal model as the PE distribution slowdown hits dealmakers where they live. 'Their bonus comp is decreasing... and as you see more continuation vehicles, that kind of payday is getting pushed out further.'" — Stephanie McAlaine, Executive Director of the ISF

Data Center Opposition Exceeds Nuclear Opposition — A Signal Investors Have Missed

It's widely understood that nuclear energy faces NIMBY opposition. What's less appreciated is that data centers are now more politically toxic than nuclear plants, fundamentally re-rating the risk profile of AI infrastructure assets.

"A May Gallup poll found that 71% of Americans oppose a data center being built near them, outpacing the 53% of Americans who oppose a new nuclear power plant in their backyard."

Two-Thirds of the US Data Center Pipeline Is Already Stalled — The Build-Out Is Not On Track

The market narrative is one of explosive AI infrastructure expansion. The operational reality is that most capacity in the pipeline isn't being built.

"Of the roughly 241 gigawatts of US data center capacity in the pipeline by the end of 2025, only about a third is under active development, according to research firm Wood Mackenzie — the rest has stalled largely due to power constraints."

3. Companies Identified

Stripe Description: Global payments infrastructure company Why mentioned: Teaming up with Advent International on a $53 billion bid for PayPal

"Stripe has teamed up with Advent International on a $53 billion bid for rival payments group PayPal."

PayPal Description: Publicly traded digital payments company Why mentioned: Target of a $53B acquisition bid from Stripe and Advent International

"Stripe has teamed up with Advent International on a $53 billion bid for rival payments group PayPal."

Chai Discovery Description: San Francisco-based AI company building models for molecular drug design Why mentioned: Raised a $400M Series C at a $3.8B valuation, led by Index Ventures, Kleiner Perkins, and Sequoia Capital — one of the largest biotech AI rounds noted

"San Francisco-based Chai Discovery, which builds AI models for molecular drug design, raised a $400 million Series C led by Index Ventures, Kleiner Perkins and Sequoia Capital at a $3.8 billion valuation."

LimX Dynamics Description: China-based humanoid robotics startup Why mentioned: Raised $200M at a $2.2B valuation; notable entrant in the fast-scaling humanoid robotics space

"LimX Dynamics, a China-based humanoid robotics startup, raised a $200 million round led by IDG Capital, Lens Technology and Stone Venture at a valuation of $2.2 billion."

TerraFirma Description: Austin-based startup building semi-autonomous construction robotics Why mentioned: Raised a $100M Series A led by Kleiner Perkins — signals institutional conviction in AI-powered physical construction

"TerraFirma, an Austin-based startup that builds semi-autonomous construction robotics, raised a $100 million Series A led by Kleiner Perkins."

PixVerse Description: Singapore-based AI video generation platform Why mentioned: Extended its Series C to $439M with Alibaba participation, signaling continued mega-round activity in generative video AI

"Singapore-based PixVerse, an AI video generation platform, extended its Series C with investment from Alibaba, Lollapalooza Capital and Ivy Capital, bringing the round's total to $439 million."

BRINC Description: Seattle-based developer of drones for public safety and emergency response Why mentioned: Raised $125M led by Motorola Solutions — strong strategic validation from a major defense/communications incumbent

"BRINC, a developer of drones for public safety and emergency responses, raised a $125 million round led by Motorola Solutions."

Singularity Description: Developer of low-cost air defense systems Why mentioned: Emerged from stealth with an $80M Series A led by Khosla Ventures and Felicis at a $400M valuation — notable defense-tech stealth launch

"Singularity, a developer of low-cost air defense systems, emerged from stealth with an $80 million Series A led by Khosla Ventures and Felicis valuing the company at $400 million."

Nous Research Description: Open-source AI agent developer Why mentioned: Raising a new round at a $1.5B valuation led by Robot Ventures — signals open-source AI agent infrastructure is attracting serious capital

"Open-source AI agent developer Nous Research is raising a new round led by Robot Ventures at a $1.5 billion valuation."

Reflection AI Description: Nvidia-backed AI model developer Why mentioned: Agreed to acquire $1B in computing power from Nebius — a notable strategic compute acquisition signaling scale ambitions

"Nvidia-backed AI model developer Reflection AI agreed to acquire computing power from Nebius in a $1 billion deal."

Uber Description: Global ride-hailing and delivery platform Why mentioned: In talks to acquire Delivery Hero in a €12B ($13.7B) deal — major consolidation move in quick-commerce

"Uber is in talks to acquire Berlin-based quick delivery company Delivery Hero in a €12 billion ($13.7 billion) deal."

Greylock Description: Storied early-stage venture capital firm Why mentioned: Launched Greylock 18, a new $1.5B fund targeting early-stage VC — signals continued confidence in early-stage investing despite market headwinds

"Greylock launched a new $1.5 billion fund, Greylock 18, targeting early-stage VC investments."

TYLsemi Description: San Jose-based developer of chiplet technology for custom silicon design Why mentioned: Emerged from stealth with a $43M round led by Matter Venture Partners — early-stage bet on custom silicon at a time of intense chip infrastructure investment

"San Jose-based TYLsemi, a developer of chiplet technology for custom silicon design, emerged from stealth with a $43 million round led by Matter Venture Partners."

4. People Identified

Harrison Rolfes Description: Senior Analyst at PitchBook Why mentioned: Provided the sharpest analytical framing of the NY data center moratorium's investment implications

"Permits and a governor's signature, it can't [be engineered around)." "A state legislature, not a chip shortage, is now setting the price."

Ted Brandt Description: Founder and CEO of Marathon Capital, an investment bank focused on energy and infrastructure Why mentioned: Contextualized the geographic risk uncertainty created by bipartisan pushback on data centers

"It's not at all clear which states are going to be data center central. And when you're starting to see New York and Texas on the same side of an issue, it's an extraordinarily weird situation."

Stephanie McAlaine Description: Executive Director of the Independent Sponsor Forum (ISF) Why mentioned: Explained the talent migration dynamic driving PE professionals toward the independent sponsor model

"Their bonus comp is decreasing. And as you see more continuation vehicles, that kind of payday is getting pushed out further."

Brett Palmer Description: President of the Small Business Investor Alliance (SBIA) Why mentioned: Highlighted the regulatory catalyst (Investing in All of America Act) boosting independent sponsor deal flow

"A lot of which will go into independent sponsor deals."

Kathy Hochul Description: Governor of New York Why mentioned: Issued the executive order creating a one-year moratorium on new large data centers — the central policy event of the issue

"New York Gov. Kathy Hochul's immediate, one-year moratorium on new data centers exceeding 50 megawatts of capacity, via an executive order on Tuesday, introduced an unexpected hurdle into the fast-growing business of building AI data centers."

5. Operating Insights

1. For AI infrastructure operators: prioritize "time-to-power" as a core competitive moat

The article reframes the AI infrastructure constraint from a technical problem (chips, interconnects) to a political one (permits, regulation). Operators who have already secured power interconnections and permits are now sitting on a strategically valuable, non-replicable asset. New entrants should model regulatory risk by state as a first-order investment variable.

"Transformers and interconnection queues: Capital can eventually engineer around [that]. Permits and a governor's signature, it can't."

2. For PE deal professionals: the independent sponsor structure is now a viable primary career track, not a fallback

With bonus compression at traditional buyout funds, continuation vehicles delaying payouts, and new SBIC leverage limits creating a larger capital pool for deal-by-deal structures, operators and dealmakers should actively evaluate whether raising a fund is the optimal path — or whether staying fundless unlocks better economics and deal selectivity.

"Both talent and capital are being nudged toward the deal-by-deal model as the PE distribution slowdown hits dealmakers where they live." "Those deals generated a median equity IRR of 23.8%, compared with 18.5% for comparable investments made by US buyouts."

3. For independent sponsors: volatility management is the critical operating discipline

The outperformance data comes with a significant dispersion warning. The upside is real (25% of deals returned 4x+), but so is the downside risk (5% total wipeouts). Portfolio construction and deal selection discipline — not just IRR chasing — will determine sustainable performance in this model.

"Investors, however, need a high tolerance for volatility. About 5% of independent sponsor deals in the study were wiped out, while a quarter returned more than 4x."

6. Overlooked Insights

1. Ivy League Credit Ratings Are Slipping — and Endowments Can't Save Them

Briefly noted but potentially significant for the higher education ecosystem, including university-affiliated research, tech transfer, and the pipeline of startup founders from elite institutions. Brown and Columbia receiving credit outlook downgrades suggests fiscal stress at institutions that produce a disproportionate share of venture-backed founders.

"Slipping credit outlooks for Brown University and Columbia University reflect monetary constraints on top-tier higher education, and endowments are ill-prepared to help."

2. European Biotech VC Is at a Record — but the US-Europe Gap Is Widening, Especially in AI

European biotech is having its best year on record, which sounds positive — but the framing is one of relative decline. The US is raising three times the capital and pulling further ahead specifically in AI-driven dealmaking, suggesting a structural divergence that could compound over time for European biotech ecosystems.

"Biotech VC is heading for a record year in Europe, but the US is still raising three times the capital and pulling further ahead in AI-driven dealmaking."