The State of the Map
The AIFI directory currently tracks 433 companies at the intersection of artificial intelligence and financial services. They collectively represent approximately $184 billion in disclosed funding across every major segment of finance and every layer of the AI technology stack.
This is the definitive overview of what that landscape looks like in 2026: where the capital has gone, which technology approaches dominate, and which parts of the map remain underbuilt.
Nine Segments, Unequal Distribution
The directory organizes companies into nine market segments based on where in finance they operate. The segments are not evenly populated.
Banking and Payments is the largest segment by funding. Stripe alone accounts for a significant portion of disclosed capital, but the segment includes dozens of companies across payment infrastructure, fraud prevention, KYC automation, and core banking modernization. The common thread is that banking generates enormous transaction volumes, and AI can extract value from that data at every step.
Trading and Markets is the most technologically mature segment. Quantitative finance adopted machine learning before the term was mainstream, and the companies in this segment — spanning systematic hedge funds, alternative data providers, and execution technology — reflect decades of investment in ML-driven alpha generation. AlphaSense, with its NLP-driven market intelligence platform, illustrates how AI is moving up the stack from raw signal to synthesized insight.
Risk and Compliance has seen the most consistent investment across market cycles. Regulatory requirements do not disappear in downturns — if anything, enforcement intensifies. Feedzai, Sardine, and their peers have built platforms that financial institutions cannot easily defund, which makes this segment structurally resilient.
Lending and Credit is where alternative data and ML scoring have had the clearest consumer impact. Upstart's model, which uses non-traditional variables to underwrite borrowers who would otherwise be declined or overpriced by FICO-based systems, demonstrated that ML can expand credit access while maintaining acceptable loss rates. The segment now includes dozens of variations on this thesis across consumer, SMB, and real estate lending.
Wealth Management, Insurance, Enterprise Finance, Crypto and Web3, and Research and Data round out the nine segments. Crypto and Web3 has seen significant funding volatility correlated with token market cycles; the companies that have survived are building infrastructure (custody, compliance, data) rather than speculative applications.
Five Layers, One Undercrowded
The directory's second classification axis is the technology layer: where in the AI stack does a company sit?
| Layer | Position | Company Count |
|---|---|---|
| Applications | 5 | ~180 |
| Automation | 4 | ~95 |
| Models | 3 | ~70 |
| Data | 2 | ~48 |
| Infrastructure | 1 | ~40 |
Applications dominate by count — most companies are building end-user products on top of existing AI capabilities. Infrastructure is the least populated layer, with approximately 40 companies. This is the classic picks-and-shovels dynamic: the infrastructure layer is expensive to build, has longer sales cycles, and often loses to large cloud providers on commodity components. The companies building durable infrastructure for financial AI — model serving, data pipelines, compliance tooling — are relatively few and relatively hard to compete with once established.
The Models layer is interesting because it sits at a strategic inflection point. General-purpose foundation models from OpenAI, Anthropic, and others have commoditized broad language capabilities. The defensible model-layer companies in finance are those with proprietary training data (earnings call transcripts, regulatory filings, market microstructure data) that produce genuinely differentiated outputs for financial use cases.
Eight AI Types: Predictive ML Still Dominates
The directory classifies companies by the primary AI technology they use. Predictive ML — supervised learning, gradient boosting, neural networks trained on labeled financial data — is by far the most common classification, appearing in roughly 200 of the 433 companies. This reflects both the maturity of the technology (predictive ML has been in production in finance since the 1990s) and its proven ROI in high-volume, structured decision contexts like fraud detection and credit scoring.
Large language models are the fastest-growing category. The shift from research curiosity to enterprise deployment happened between 2023 and 2025, and the companies building LLM-native financial products are still early in their growth curves. The primary use cases are document processing, research synthesis, compliance drafting, and client communication — all contexts where the value of natural language understanding is clear and the cost of hallucination can be managed through retrieval-augmented approaches and human review.
Agentic AI is the newest category, with 35 companies now classified as primarily agentic builders. The distinction from predictive ML and LLM applications is autonomy: agentic systems take multi-step actions without human approval at each step, and they adapt based on feedback from their environment. Gartner's projection that AI agents will autonomously make 15 percent of work decisions by 2028 looks plausible from current trajectories in financial services specifically.
Graph analytics, computer vision, reinforcement learning, and infrastructure-focused AI make up the remainder.
Structural Patterns Worth Noting
The acquisition rate is approximately 10 percent. Roughly 43 of the 433 companies in the directory have been acquired. The acquirers are predictable — large banks, major fintech platforms, and enterprise software companies looking to buy rather than build AI capabilities. The acquisition targets are typically companies with defensible data assets or proprietary models rather than pure application-layer products.
Geographic concentration is significant. The Americas — primarily the United States — accounts for the majority of companies and the overwhelming majority of funding. EMEA has meaningful representation in compliance technology, where European regulatory requirements create natural demand for local expertise. APAC is underrepresented relative to the size of the addressable market, though this is shifting with the growth of embedded finance in Southeast Asia and AI investment in Japan and South Korea.
Funding stage distribution is skewed toward early. The majority of companies in the directory are at seed or early stage, reflecting the recent acceleration of AI investment in finance. Late-stage and public companies are fewer but account for a disproportionate share of total disclosed funding — the distribution follows a power law typical of venture-backed technology markets.
What the Map Does Not Show
The AIFI directory tracks companies that have raised disclosed funding, have a meaningful web presence, and are operating in finance-adjacent AI markets. It systematically undercounts stealth-mode companies, large bank internal AI teams, and consulting firms that have built significant AI practices without raising external funding.
The on-chain agent ecosystem — 144 ERC-8004-registered agents currently classified as finance-relevant — represents a parallel layer not captured in the company directory at all. These agents have no funding rounds, no headquarters, and no traditional business model. They are infrastructure objects with on-chain identities. As this layer matures, it will increasingly interact with the companies in the directory in ways that do not map cleanly to existing competitive analysis frameworks.
Navigating the Map
The full interactive market matrix — showing company density at every segment-layer intersection — is available at aifimap.com/directory. Individual company profiles include funding data, team information, technology classification, and related companies by segment and layer.
The patterns visible in the map today — infrastructure undercrowded, predictive ML dominant, banking funded most heavily, agentic AI accelerating — are the structural facts that will shape where the next wave of investment and competition concentrates. The companies that understand these patterns have a better-informed basis for building, investing, and competing in this market.