Seligman Investments has doubled the deployable capital available to Seligman Ventures to $1 billion, expanding its ability to invest across the physical infrastructure supporting artificial intelligence. The increase comes less than a year after the venture operation launched in February 2026 with $500 million in deployable capital. Since then, Seligman Ventures has invested more than $300 million across 14 investments spanning AI hardware, connectivity and cybersecurity. The move gives the Silicon Valley technology investor substantially more capacity to pursue companies positioned around the infrastructure constraints emerging as AI systems expand.

Seligman Expands Its AI Infrastructure Bet

The additional capital reflects a broader shift in technology investment as computing infrastructure returns to the center of the AI growth story. Seligman sees opportunities across accelerators, networking equipment, power systems and cooling technologies, areas that increasingly determine how quickly new AI computing capacity can reach operation. The firm is also continuing to evaluate cybersecurity businesses alongside companies supplying physical AI infrastructure. Its strategy effectively places capital behind several layers of the technology stack rather than concentrating solely on AI software businesses.

Investor interest in hardware has strengthened as the AI buildout demands more chips, connectivity and supporting infrastructure. Venture investment in U.S. and Canadian startups reached a record $392 billion during the first half of 2026, according to Crunchbase data cited by Reuters. Semiconductor startups accounted for about $10.7 billion of that investment and were on pace to exceed their total from the previous year. The figures illustrate how capital is moving toward technologies needed to support increasingly infrastructure-intensive computing environments.

“The amount of deal flow we have gotten is 10 or 15 times more than what I anticipated when I joined,” Umesh Padval, managing partner at Seligman Ventures, told Reuters. “We didn’t want to lose the AI, cybersecurity or the datacenter train.” The comments point to the volume of investment opportunities Seligman says it has encountered since launching its venture operation. They also explain why the firm has increased available capital so soon after establishing the platform.

Physical AI Constraints Create New Investment Targets

AI infrastructure has become a larger investment category because expanding compute requires much more than access to advanced processors. High-performance systems depend on networking capacity to connect accelerators, electrical infrastructure capable of supporting large computing loads and cooling architectures that can remove the heat produced by dense hardware. Seligman’s investment mandate places those physical dependencies alongside semiconductors as potential areas for venture deployment. That approach gives the firm exposure to technologies that could influence how efficiently developers and infrastructure operators expand AI capacity.

The strategy is particularly relevant as the industry works to translate demand for AI computing into functioning physical capacity. Accelerator availability represents only one part of that process because networking, power delivery and thermal management must also support the hardware once it reaches a facility. Investment opportunities can therefore emerge around components that improve the movement of data, electricity and heat through computing environments. Seligman’s expanded capital pool provides additional capacity to invest across those infrastructure layers as new companies attempt to address them.

Seligman’s existing portfolio already reaches into technologies used within data center environments. Reuters identified AI computing company SambaNova, which competes with Nvidia in accelerated computing, and optical technology company Lumilens among the firm’s investments. The portfolio demonstrates how Seligman is spreading its exposure across computing and connectivity technologies rather than treating AI infrastructure as a single hardware category. That breadth could become increasingly important as infrastructure requirements evolve alongside successive generations of AI systems.

Seligman Uses a Barbell Investment Strategy

Seligman Ventures is pursuing what Padval described as a barbell strategy that combines early-stage investments with late-stage and pre-IPO opportunities. This structure allows the firm to invest in younger infrastructure companies while maintaining exposure to businesses approaching larger commercial scale or public markets. Meanwhile, the expanded $1 billion capital base gives the venture operation more room to participate across those different stages. Seligman expects accelerators, networking, power, cooling and cybersecurity to remain areas where it will search for new investments.

The strategy also connects private technology investing with Seligman’s established public-market research operation. Paul Wick leads the public-markets business and manages the roughly $29 billion Columbia Seligman Technology and Information Fund, while the firm also oversees a $7.5 billion technology hedge fund business. Seligman says its venture team combines private-company sourcing and board involvement with analysis from its public-market operation. That model gives investment teams another way to examine how emerging technology companies could compete with established businesses.

“We have a private knowledge of 400 companies in the private world. We marry that with public-market data analysis to bring out a thesis as to which ones we invest in,” Padval said. He expects Seligman Ventures to make five to 10 new investments during the next 12 months. The larger capital allocation gives the firm room to maintain that pace while participating in companies at different stages of development.

Board Participation Shapes Seligman’s Venture Model

Seligman is also taking a more active governance role in its private-company investments than many public-market investors that enter venture deals. The firm currently holds six board seats and three board observer positions across its private investments. That involvement gives Seligman closer access to portfolio companies as they develop products, compete for customers and prepare for later financing or potential public listings. The approach differentiates its venture operation from crossover strategies centered primarily on obtaining financial exposure to private technology businesses.

“This time around we have two people who are going on boards of companies,” Wick said. “Our ability to protect ourselves and to better understand our private-company investments is just so enhanced.” His comments reference lessons from an earlier venture effort launched in 1997 that later ended following the dot-com collapse. Wick cited limited board representation, dependence on third-party deal sourcing and team selection among the lessons Seligman took from that experience.

The renewed venture push arrives as technology companies increasingly remain private for longer periods, changing how public-market investors access high-growth businesses. Seligman’s model gives its investment teams an earlier position in companies that might otherwise remain outside public portfolios for years. However, the firm is coupling that access with board participation and its existing technology research capabilities rather than relying solely on later-stage private financing rounds. That structure becomes particularly relevant in AI infrastructure, where companies can require substantial capital before their technologies achieve broad commercial deployment.

AI Infrastructure Moves Deeper Into Venture Capital

The increase from $500 million to $1 billion signals that Seligman expects infrastructure constraints to remain investable opportunities as AI deployment grows. Chips attract much of the attention around AI computing, but the systems surrounding those processors increasingly determine whether additional capacity can operate at the required scale. Networking equipment must connect larger clusters, power infrastructure must support expanding electrical demand and cooling systems must manage the resulting thermal loads. Capital directed toward those supporting technologies could therefore influence how quickly the next generation of AI infrastructure reaches commercial operation.

For AI infrastructure operators, the investment shift also expands the potential funding ecosystem around technologies that sit outside the accelerator itself. Startups developing connectivity, power or cooling products often address constraints that become more visible as computing density and cluster scale increase. A larger pool of venture capital does not remove those engineering challenges, but it can give companies working on them additional resources to develop products and pursue commercial deployments. Seligman’s $1 billion commitment places the firm among investors explicitly positioning around that opportunity.

The timing is notable because AI investment is increasingly intersecting with physical infrastructure decisions rather than remaining concentrated in software development. Semiconductor availability, network architecture, electrical capacity and thermal systems now form interconnected parts of the same deployment equation. Investors that can evaluate those dependencies may gain exposure to businesses supplying critical components before infrastructure demand reaches its next stage. Seligman is positioning its expanded venture operation around that premise as it prepares to make additional investments over the coming year.

Seligman Targets the Infrastructure Behind AI Growth

Seligman’s decision to double deployable venture capital less than a year after launch shows how quickly AI infrastructure has moved into the investment conversation. The firm has already committed more than $300 million across 14 investments, while its expanded mandate leaves room for further positions across accelerators, networking, power, cooling and cybersecurity. Its combination of early-stage, late-stage and pre-IPO investing also allows it to follow infrastructure businesses through different phases of their development.

The broader implication extends beyond one investment platform because the AI buildout increasingly depends on solving physical constraints alongside advances in computing hardware. More capable processors create value only when surrounding networks, electrical systems and thermal infrastructure can support their deployment. Seligman’s expanded capital pool represents a bet that companies solving those supporting challenges can become significant technology businesses themselves. As AI infrastructure spending broadens, the technologies behind the compute may become as important to investors as the processors receiving most of the attention.