Union Square Ventures in New York Announces New Fundraising of $900 Million
New York venture capital firm Union Square Ventures has announced a new fundraising of $900 million, specifically targeting startups in large markets being reshaped by technology. The funds are divided into two parts: a $500 million early-stage core fund and a $400 million opportunity fund for later follow-on investments and additional investments in existing portfolios.
The previous core fund was $275 million in 2024, making this early-stage fund nearly double; the comparable previous opportunity fund was $350 million, with the new opportunity fund increasing to $400 million. The firm stated that as product development becomes easier, high-quality projects require larger rounds and higher prices, necessitating more capital for leading investments and longer-term support. The number of investments will remain roughly the same as before, but the size of individual investments is expected to increase according to the founders' projections for 2026, explicitly covering capital-intensive sectors such as robotics, manufacturing, and energy.
The active general partners have increased to four: Fred Wilson, Rebecca Kaden, Michael Mignano, and Nick Grossman. Michael Mignano, co-founder of Anchor, is the new general partner; Jared Hecht (co-founder of GroupMe and Fundera) is appointed as a venture capital partner; Scott Belsky (co-founder of Behance and former Chief Product Officer at Adobe) is the product advisory partner. Andy Weissman, Albert Wenger, John Buttrick, and Brad Burnham will continue to support the firm and portfolio companies but will no longer be part of the core partnership.
The investment thesis has four points: application-layer companies are reshaping markets with artificial intelligence rather than just accelerating old processes; investing in "edge data" to obtain previously unreachable offline data through physical intelligence, robotics, manufacturing, and sensor networks; betting on intelligent products that align user incentives with trust, security, memory, and cost; and merging the energy fund established in 2021 into the core fund, as all these directions rely on more energy.
The firm compares this increase in funding to its judgment in 2004 when Fred Wilson and Brad Burnham bet on early companies in internet infrastructure, subsequently entering early investments in Twitter, Twilio, Etsy, MongoDB, and Coinbase. The written statement has shifted to betting on "intelligence," stating that the infrastructure for artificial intelligence has been laid out, but defining applications are yet to emerge.
Limited partners are giving money to Union Square Ventures, which will then use larger checks to lead early rounds and use the opportunity fund for later investments. The sellers are founding teams needing computing power, hardware, and energy capital expenditures, while the buyers are established early funds that still want to maintain thesis-driven investments but must keep up with round pricing. The beneficiaries are companies in robotics, manufacturing, energy, and edge data, as well as institutions that can provide leading investment amounts; the pressured parties are early funds with smaller checks and existing investors whose returns are compressed due to inflated valuations.
ABAB AI Insight
Union Square Ventures' historical path has been characterized by small funds, early rounds, and network effects. The first fund in 2004 was approximately $125 million, leading a $5 million Series A round for Twitter in 2007, followed by investments in Etsy; in 2013, it led a $5 million Series A round for Coinbase at a share price of about $0.20, which was worth billions at the time of its IPO. This same approach was exposed to downturns during the 2022 crypto winter: Coinbase's valuation significantly dropped, yet Fred Wilson publicly maintained exposure to crypto, indicating that the firm prefers to endure cycles rather than easily change its thesis.
This $900 million is not a scattergun approach but rather consolidates limited partners' money into two tools: a $500 million early fund responsible for leading investments and a $400 million opportunity fund for holding validated positions longer. The energy fund established in 2021 has been merged back into the core fund, effectively making electricity, robotics, and manufacturing the main strategy rather than a side pocket. The motivation is specific: Series A rounds have now reached about $10 million, and AI companies need to burn cash for computing power and customer acquisition. Continuing to use the old core fund of $275 million would mean losing leading and follow-on rights to larger platform funds.
In contrast, Andreessen Horowitz has continuously scaled its funds in crypto and AI, while Benchmark has long refused to expand its scale. Union Square Ventures has historically been closer to the latter, but this round shifts towards the former's larger checks while retaining the latter's number of investments, transitioning from "constraint-based early" to "selective platform," rather than becoming a late-stage giant. Its industry position is an early foothold in application layers and physical intelligence, not a capital arms race for model training.
Structurally, this represents a transfer of pricing power, accompanied by capital concentration. AI has lowered the cost of product development while increasing the round prices for winning companies, shifting the investment rights from "who understands first" to "who can still lead at higher prices and follow through completely." The merger of the energy fund into the core fund is because the marginal costs of models, robotics, and sensors ultimately fall on electricity and hardware, thus shifting capital from pure software equity to a bundled configuration of software, physical systems, and energy.
ABAB News · Cognitive Laws
- Making products has become cheaper, but buying winners has become more expensive.
- Small funds maintain judgment, while large checks secure positions.
- The endpoint of software often lies in electricity and hardware.