Two reports published within weeks of each other quantify what the AI industry has been feeling since late 2024: the money is moving. Not shrinking, not plateauing, but redirecting at scale from generative AI experiments toward the infrastructure and vertical deployments that make autonomous agents production-ready.

BCC Research published its State of the AIT Industry Q1 2026 review on May 14, reporting that global venture funding in AI technology reached approximately $297 billion in 2024, with 80 to 81 percent directed toward AI-focused companies rather than traditional tech conglomerates. Stanford’s 2026 AI Index, released in April, tracked an even larger number for 2025: $344.7 billion in private AI investment, a 127.5 percent increase from 2024, according to Digital Information World’s summary of the findings. Total corporate AI investment across all sources hit $581.7 billion in 2025, up 130 percent year over year.

These are not projections. They are settlement figures.

The Scale of the Shift

The raw numbers require context. The BCC Research figure of $297 billion represents venture funding specifically. Stanford’s $344.7 billion captures all private AI investment, a broader category that includes corporate rounds, private equity, and late-stage financing alongside venture capital. Both track the same trajectory: capital doubling on an annual basis.

According to the BCC report, the first half of 2025 saw total AI deal value increase by over 120 percent compared to H1 2024. U.S. private AI investments reached $109.1 billion in 2024, nearly 12 times higher than China’s $9.3 billion, according to BCC Research. Stanford’s AI Index extends that gap further: U.S. AI investment in 2025 reached $285.9 billion, 23.1 times greater than the next-highest country, China, at $12.4 billion, according to Digital Information World.

The corporate venture arms tell their own story. Google, Microsoft, and Amazon collectively invested more than $50 billion in AI in 2025, per BCC Research. Meta invested over $14 billion in AI infrastructure and tooling.

Where the Money Is Going

The headline numbers mask the more significant story: what the capital is buying. The BCC report identifies a “fundamental shift as organizations abandon pilot programs for enterprise-wide AI deployment, driving near-doubling of enterprise spending year-over-year.” The emerging technology focus has moved from large language models toward agentic AI for complex multistep tasks and quantum machine learning.

New Market Pitch, which tracks pure-play agentic AI companies exclusively, puts the narrower number in sharper focus. Between 2022 and 2025, agentic AI startups raised a combined $4.4 billion across 101 equity deals. Full-year disclosed funding rose from approximately $1.5 billion across 31 deals in 2024 to $2.9 billion across 50 deals in 2025. January through April 2026 already produced approximately $1.1 billion across 29 deals, roughly twice the capital and more than three times the deal activity of the comparable period in 2025.

The composition of that spending reveals where conviction is concentrating. Vertical AI agents, meaning agents built for specific industries like cybersecurity, healthcare, procurement, and compliance, account for 48.3 percent of 2026 year-to-date deals and 54.6 percent of capital, according to New Market Pitch. Agent execution infrastructure, covering runtimes, sandboxes, identity, observability, and security testing, captured 20.7 percent of deals. Agent development platforms rebounded in early 2026 after a weaker 2025, reaching approximately $124 million across five deals by May.

The takeaway: investors are not funding “AI” as a category. They are funding the specific layers that make agents deployable, governable, and measurable in production.

The Enterprise Deployment Signal

The reports align on one critical observation: enterprises are done experimenting.

JPMorgan Chase now spends $2 billion annually on AI specifically, per BCC Research, within a broader technology budget that InfotechLead reports at $17.5 billion. The bank is rolling out agentic AI systems executing end-to-end tasks across payments, customer service, and back-office functions. Bank of America spent approximately $13 billion on technology in 2025, with plans to increase tech development spending by 10 percent in 2026, according to Business Insider.

NVIDIA and Eli Lilly announced up to $1 billion for an AI co-innovation lab. The European Commission launched $356 million in Horizon Europe AI funding calls. These are not proof-of-concept budgets. They are operational commitments from institutions that have moved past the “should we adopt AI?” question entirely.

The BCC report frames the convergence precisely: “enterprise demand and aggressive capital deployment from private and public markets are converging to accelerate production-scale deployment.”

The Concentration Problem

The aggregate numbers conceal a distribution that should concern smaller players and investors alike. According to New Market Pitch, the top three agentic AI deals in 2026 year-to-date captured 44 percent of all capital. The top 10 captured 78 percent. The bottom half of deals captured only 11.5 percent.

The average round in early 2026 is $36 million; the median is $19 million. That gap confirms headline funding totals overstate the financing environment for a typical agentic AI company. March 2026 alone contributed approximately $650 million, more than 60 percent of year-to-date capital, driven by a few large security, customer-service, and agent-control-plane rounds.

North America dominates. It captured approximately $863 million across 23 deals in 2026 year-to-date, about 82 percent of capital and 79 percent of deals. Europe remains active but undercapitalized. Asia-Pacific is episodic.

The honest read: the agentic AI market is not a rising tide lifting all boats. It is a selective funnel where capital concentrates around measurable workflow replacement, security infrastructure, and enterprise control points. Generic agent wrappers and broad autonomy claims without a clear buyer are getting cut.

The Headwinds

Both reports flag risks that the investment euphoria tends to obscure. BCC Research notes that “deployment timelines face headwinds from power shortages and supply chain limitations despite record spending levels.” Data quality issues, talent shortages, and unclear ROI continue to hinder scaled implementation, with “high failure rates and project abandonment creating a challenging operational environment.”

IEEE Spectrum, reporting on the Stanford AI Index, highlights a separate pressure: AI compute capacity has grown 3.3 times annually since 2022, a 30-fold increase since 2021. Training frontier models now generates enormous carbon emissions, with xAI’s Grok 4 estimated at over 72,000 tons of CO2-equivalent, up from 5,184 tons for GPT-4. Local governments in the United States are beginning to restrict or ban new data center development entirely.

The capital is flowing faster than the infrastructure can absorb it.

What This Capital Is Actually Buying

Strip away the aggregate figures and the investment thesis becomes legible. The 2024-2025 wave funded models. The 2025-2026 wave is funding the operating environment around those models: execution infrastructure, governance layers, vertical specialization, and the testing and observability tooling that makes autonomous agents safe enough for enterprise procurement officers to sign off on.

This is the pattern every major technology wave follows. First the core technology gets funded. Then the surrounding infrastructure gets funded. Then the verticals get funded. AI agents entered the infrastructure and vertical phase in late 2025, and the capital curves confirm it.

The $297 billion BCC figure and the $344.7 billion Stanford figure are measuring different slices of the same shift. The agentic-specific $4.4 billion from New Market Pitch is the leading edge. The question is no longer whether agents will be deployed at scale. The capital markets settled that. The question is which infrastructure layers capture the margin, and which agent companies end up in the 78 percent or the 11.5 percent.