Venture capital flowing into fintech startups dropped 21.4% year over year in the third quarter of 2026, while the number of deals fell nearly 31%, according to Crunchbase data published in early October. The figures point to a narrowing funding landscape where fewer companies are raising money and large transactions carry more weight. Investors are drawing sharper lines between startups with AI capabilities or market leadership and those without, the report finds.
Fintech companies worldwide secured $10.6 billion in the third quarter of 2026, down 28.4% from $14.9 billion in Q2 2026 and 21.4% from $13.5 billion in the same quarter a year earlier, according to the report. The number of funding rounds hit its lowest point since the first quarter of 2018, with just 687 deals recorded globally—a 12.9% decline from the previous quarter's 789 rounds and a 31% drop from 995 rounds in Q3 2025. Funding from rounds of at least $100 million fell 43.3% quarter over quarter, from $9.07 billion to $5.15 billion, accounting for roughly 93% of the overall dollar decline. By contrast, rounds below $100 million declined just 5.1% sequentially. Year over year, large rounds showed more resilience, falling only 7%, while smaller rounds dropped 31.4%. Early-stage fintech funding reached $4.23 billion, up 11% from the previous quarter and down just 0.6% year over year, but one transaction—Ant International's $1.2 billion Series A—accounted for 28.4% of that total. Without that deal, early-stage funding would have fallen 20.4% sequentially and 28.8% year over year. Angel and seed funding totaled $1.12 billion, down 19.7% from Q2 and 35.2% from Q3 2025, while late-stage funding fell to $4.8 billion, down 40.4% quarter over quarter and 17.3% year over year.
The report describes a bifurcated market where capital concentrates in fewer, larger rounds, with infrastructure, payments, and stablecoin rails drawing investor interest supported by greater regulatory clarity. QED Investors partner Laura Bock said "the market remains bifurcated—AI vs. not and market leaders vs. not." Forerunner founder Kirsten Green noted that for Series A and B rounds, "evidence counts for than narrative," with investors demanding stronger proof of performance, efficient growth, a credible path to profitability, and lasting customer demand before writing checks. The report also highlights geographic concentration: the United States captured $5.38 billion, or 50.5% of global fintech funding in Q3, followed by Singapore with $1.28 billion (12.1%), heavily driven by Ant International's round. Despite U.S. dominance overall, the quarter's five largest funding rounds went to companies outside the United States, including Ant International, Saudi Arabia's Barq ($329.5 million Series A), Hong Kong's Qupital ($300 million Series C), Indonesia's Ajaib ($270 million Series C), and Saudi Arabia's Tabby ($233 million Series F).
The report finds that AI is reshaping investor priorities, primarily by redirecting investment within fintech rather than creating entirely new capital pools, though agents and payments are attracting some fresh funding. Companies without clear AI offerings can still raise money, but investors increasingly expect evidence that the technology improves economics, speed, or customer reach. Bock said the distinction matters between companies adding AI to existing processes and those building products around it from the ground up—the strongest AI products deliver substantial improvements that drive rapid adoption and willingness to pay. She noted that when investors call a company "AI," they usually mean its growth rate is high, making growth the proxy. Fundraising expectations are rising as AI makes products faster to build and launch, with early revenue carrying less weight than sustained growth; while $1 million in annual recurring revenue once served as a common Series A benchmark, investors now want greater traction and stronger momentum, including revenue that has at least tripled over six months. "Efficiency still matters, but growth is back in focus," Bock said, adding that expectations of larger potential outcomes are pushing valuations higher. A small group of companies with clear momentum command aggressive valuations, while others face greater discipline—Green pointed to Ramp and Monzo as similarly sized businesses with widely different valuations, attributing Ramp's premium to investor enthusiasm for its AI momentum and growth trajectory.
The report suggests AI could reshape financial services by sharply reducing the costs of underwriting, compliance, and customer service while enabling more personalized financial products. QED is particularly interested in advice and wealth management, underwriting, insurance claims, risk and compliance, and finance back-office operations, alongside stablecoins and tokenization, with Bock stating "the financial services operating system will be rebuilt end-to-end." Green said payments initiated by AI agents have moved "from 'if' to 'when'" in her conversations, as major networks and processors introduced protocols to support them, and she's seeing more fintech companies seek ownership of regulated infrastructure rather than rely on outside providers. However, she sees potential overfunding in stablecoin infrastructure, where many startups are pursuing the same customers, and in tools that let AI agents complete purchases for buyers, where startup formation has outpaced transaction volume. Green sees more opportunity in the systems that make agent-driven commerce trustworthy, including fraud prevention, identity verification, agent permissions, and tools that help merchants accept those purchases, arguing that payment infrastructure is likely to become commoditized, leaving more value with companies that own customer relationships and earn trust. The bifurcation between AI-native companies and traditional fintech operators suggests that technical capabilities alone won't determine which startups secure funding—instead, the ability to demonstrate that AI translates into measurable business advantages will separate winners from those left behind.