For the past two years, investors seeking exposure to artificial intelligence have largely chased the companies building the models: OpenAI, Anthropic, xAI, and other headline-grabbing startups promising to reshape the economy.
But a growing body of venture data suggests the real AI trade may have quietly shifted elsewhere. According to PitchBook’s latest Infrastructure SaaS VC Trends report, capital is increasingly flowing into the software and systems that make AI possible rather than the applications consumers see.
Infrastructure SaaS deal value reached $17.1 billion across 218 deals during the first quarter of 2026, representing a 114.2% increase in value quarter-over-quarter and a 40.6% jump in deal count, the report noted.
At first glance, skeptics could dismiss those figures as being heavily skewed by Databricks’ massive $7 billion financing round completed in February. But even after removing that megadeal from the equation, infrastructure SaaS funding still climbed 26.5% quarter over quarter, showing “broad demand for the software layers supporting AI workloads,” the report wrote.
This behavior suggests that investor appetite for AI is broadening beyond a handful of elite foundation model companies and spreading across the layers of software required to train, deploy, secure and manage increasingly complex AI systems.
Where the Money Is Actually Going
“The diversification demonstrates investors are not crowding into a single theme,” PitchBook analysts wrote in the report. “Funding is accruing across the full stack on which AI workloads run, from development tooling through data architecture to operational control.”
The data shows just how widespread that shift has become.
Excluding Databricks, application infrastructure led the quarter with roughly $2.7 billion invested across 54 deals. DevOps followed with $2.2 billion across 40 transactions, while data software and systems attracted $1.5 billion across 27 deals. ITOps accounted for another $870.7 million across 33 deals.
“As anticipated, we continue to see broad infrastructure investment due to the ever-evolving demands of AI models and applications on the landscape of infrastructure solutions,” the report stated.
As enterprises move from pilot programs to production deployments, their needs evolve. Organizations suddenly require data governance tools, observability platforms, automated testing systems, cloud optimization software and developer infrastructure capable of handling enormous AI workloads.
Those demands are creating entirely new winners.
PitchBook noted that investment strength extended across the infrastructure stack, reflecting “robust underlying strength in the broader asset class” and serving as a “counterpoint to the concentrated capital deployment characteristic of the AI startup ecosystem.”
Exit markets are reinforcing the trend. Infrastructure SaaS generated a record $40.5 billion in exit value during the quarter, driven largely by Alphabet’s $32 billion acquisition of cybersecurity startup Wiz and Marvell Technology’s $6 billion acquisition of Celestial AI.
While the number of exits remained relatively subdued, the size and strategic importance of those transactions suggest major buyers are willing to pay a premium for critical infrastructure assets tied to the AI ecosystem.
The implications for investors are significant. “The IPO pipeline behind Databricks, Stripe and Rippling suggests the public-market reopening could come into clearer view over the next 12 months. The vendors positioned to benefit are the ones owning the layers AI quietly depends on, rather than simple AI-powered marketing,” the report concluded.
Photo: Shutterstock