New York’s AI boom is no longer a side story to Silicon Valley.
In 2026, investors are writing some of the largest technology checks New York has ever seen. But the most important part of the story is not simply how much money is coming into the city. It is where that money is going.
The biggest checks are not being spread evenly across every startup that puts “AI” on its website. Capital is concentrating around a smaller group of companies that have expensive technology to build, large enterprise customers to serve, difficult industries to understand, or some combination of all three.
That pattern matters.
It suggests New York is developing a very different AI economy from the one many people associate with San Francisco. New York certainly has companies building models, infrastructure, developer tools, and generative media. But an unusually large part of the city’s AI ecosystem is tied to industries already deeply rooted here: finance, professional services, media, real estate, healthcare, security, compliance, accounting, and enterprise software.
To understand that shift properly, NYC Tech Journal built an original dataset of the most heavily funded private AI companies connected to New York City.
Our analysis identified 21 companies at the funding cutoff, rather than 20, because two companies were effectively tied at the bottom of the ranking. Together, these businesses have raised a conservative $11.58 billion or more in disclosed funding.
More importantly, we estimate that companies in this group received at least $4.41 billion in confirmed primary venture equity during 2026 alone, using a deliberately conservative methodology.
That gives us a much clearer picture of what investors actually believe will matter in the next phase of AI.
And the answer is not “another chatbot.”
Original Research: How NYC Tech Journal Built This AI Funding Dataset
Startup funding numbers are surprisingly difficult to compare.
One database may count debt while another counts only equity. A company may announce “$1 billion in financing,” even though part of that amount involves existing shareholders selling shares rather than new money entering the business. News reports may also describe a funding round that is being negotiated even though it has not actually closed.
We wanted the dataset behind this article to be useful rather than simply impressive-looking.
Our inclusion rules
For this analysis, a company had to meet four main conditions.
First, it had to be a private company with substantial headquarters or primary operating roots in New York City.
Second, artificial intelligence had to be central to its product, infrastructure, or business model. Simply using machine learning somewhere inside an ordinary software product was not enough.
Third, the company had to remain an independent private business at our August 29, 2026 research cutoff.
Fourth, we needed enough reliable public information to estimate cumulative funding.
This approach means the dataset includes everything from AI infrastructure companies such as VAST Data and Modal to specialized AI businesses such as Rogo, Basis, Norm AI, EliseAI, Profound, and Runway.
Cyera is included because its AI-based classification technology and newer “trust layer for enterprise AI” strategy place AI at the center of its platform rather than treating it as a small add-on. Cyera itself describes AI as foundational to how its system understands sensitive enterprise data.
How we treated unusual funding rounds
For cumulative funding, we used publicly reported capital raised, drawing primarily from company announcements and established private-market databases.
For our separate analysis of 2026 venture capital flow, we were stricter.
Known debt was removed. ASAPP’s $24.1 million 2026 debt financing, for example, remains relevant to the company’s funding history but is not counted in our strict 2026 venture-equity calculation.
Mirage’s $75 million General Catalyst Customer Value Fund financing is also shown in the company discussion but excluded from our strict equity total because it is structured differently from a conventional venture round.
VAST Data requires another adjustment. Its 2026 Series F transaction was roughly $1 billion when primary and secondary capital were combined, but reports indicate that a little over half went to existing shareholders rather than directly into the company. We therefore use approximately $500 million of primary capital when estimating 2026 new venture dollars.
Finally, we did not add EliseAI’s reported $300 million August 2026 financing to the 2026 total. As of our cutoff, credible reports still described the deal as being negotiated rather than definitively closed.
That decision is important. Adding every reported deal before it closes makes startup ecosystems look larger than they really are.
The Most-Funded AI Startups in NYC
Here is the resulting NYC Tech Journal dataset.
| Rank | Company | Approx. disclosed funding | Core AI market | Major 2026 financing included in strict VC analysis |
|---|---|---|---|---|
| 1 | Cyera | $2.3B+ | AI security and governance | $1.0B across two rounds |
| 2 | AlphaSense | $1.747B | Enterprise and financial intelligence | $350M |
| 3 | VAST Data | ~$1.4B | AI data infrastructure | ~$500M primary |
| 4 | Dataiku | ~$865M | Enterprise AI platform | — |
| 5 | Runway | ~$860M | Generative video and world models | $315M |
| 6 | Flourish | $500M | Brain-inspired frontier AI | $500M |
| 7 | Modal | ~$466M | AI cloud infrastructure | $355M |
| 8 | General Intuition | ~$454M | Action models and physical AI | $320M |
| 9 | EliseAI | ~$422M confirmed | Housing and healthcare AI | Pending 2026 round excluded |
| 10 | ASAPP | ~$337M | Contact-center AI | Debt excluded |
| 11 | Rogo | ~$311M | AI for finance | $235M |
| 12 | Standard Bots | ~$287M | AI-native robotics | $200M |
| 13 | Norm AI | $260M+ | Legal and compliance AI | $120M |
| 14 | Cadence | ~$241M | Clinical AI | $100M |
| 15 | Taktile | ~$213M | AI financial decisioning | $110M |
| 16 | Mirage | $175M+ | AI video | Financing excluded from strict equity calculation |
| 17 | Hebbia | ~$161M | AI for knowledge work and finance | — |
| 18 | Profound | $155M+ | AI search marketing | $96M |
| 19 | OpenRouter | ~$153M | Model routing and AI infrastructure | $113M |
| 20= | Pinecone | ~$138M | Vector database infrastructure | — |
| 20= | Basis | ~$138M | AI accounting agents | $100M |
Company totals are drawn from public company announcements and private-market data. Cyera has raised more than $2.3 billion; AlphaSense about $1.747 billion; VAST Data roughly $1.4 billion; Dataiku about $865 million; and Runway about $860 million.
Modal has raised roughly $466 million, while General Intuition reached approximately $454 million after its $320 million 2026 Series A. EliseAI’s confirmed financing is roughly $422 million, although a further $300 million round is under discussion.
Rogo has raised about $310.5 million, Standard Bots about $287 million, Norm AI more than $260 million, Cadence $241 million, and Taktile about $213 million.
At the lower end of our cutoff, Mirage reports more than $175 million, Hebbia about $161.1 million, Profound more than $155 million, OpenRouter around $153 million, Pinecone $138 million, and Basis about $138 million.
What the Dataset Tells Us Immediately
The first big finding is concentration.
These 21 companies have raised roughly $11.58 billion combined, but the money is not evenly distributed.
Cyera, AlphaSense, and VAST Data alone account for about 47% of all cumulative capital in the dataset.
Add Dataiku and Runway, and the top five account for roughly 62%.
The top 10 account for just over 80%.
Chart: How concentrated is NYC AI funding?
| Group | Share of cumulative funding |
|---|---|
| Top 3 companies | 47.0% |
| Top 5 | 61.9% |
| Top 10 | 80.7% |
| Remaining 11 | 19.3% |
Capital concentration
Top 3 ███████████████████████ 47%
Top 5 ███████████████████████████████ 62%
Top 10 ████████████████████████████████████████ 81%
This tells us something important about the AI market.
There may be hundreds of companies building AI products in New York, but venture dollars are being distributed according to a power law. A relatively small number of businesses absorb a very large percentage of total investment.
The median company in our dataset has raised about $310.5 million.
The average, however, is approximately $551.5 million.
That large gap between the median and average exists because giant rounds at companies such as Cyera, AlphaSense, and VAST Data pull the average sharply upward.
For founders, this is a useful reminder. Headlines about billion-dollar AI funding rounds do not describe the normal fundraising experience, even among successful AI companies.
At Least $4.41 Billion Has Flowed Into This Cohort in 2026
Cumulative funding tells us which businesses have built the largest capital bases.
But to understand where venture capital is going now, we need to isolate 2026.
Using only confirmed primary equity or conventional venture rounds that we could reasonably compare, our dataset produces a conservative 2026 total of approximately $4.41 billion.
That excludes known debt, excludes Mirage’s alternative growth financing, excludes undisclosed rounds, excludes pending EliseAI financing, and counts only about $500 million of VAST Data’s roughly $1 billion mixed primary-secondary transaction.
Even with those restrictions, the number is huge.
AlleyWatch estimated total NYC startup funding at approximately $17.41 billion through July 2026.
That means this small group of major AI companies alone received capital equal to roughly 25% of all New York City venture funding through July, although the figures should be treated as approximate because our methodology and AlleyWatch’s broader market methodology are not identical.
One quarter of the city’s startup dollars flowing into a handful of large AI companies is not a small trend.
It is a structural shift.
Where the 2026 Money Is Actually Going
We grouped the companies by their primary economic role rather than simply calling all of them “AI software.”
That gives us a much more useful view.
Chart: Confirmed 2026 primary VC by AI category
| Category | Approx. 2026 capital | Share of analyzed capital |
|---|---|---|
| Frontier AI and robotics | $1.02B | 23.1% |
| AI security and governance | $1.00B | 22.7% |
| AI infrastructure and platforms | $968M | 21.9% |
| Finance, legal and professional workflows | $915M | 20.7% |
| Generative media | $315M | 7.1% |
| Vertical operations and enterprise agents | $196M | 4.4% |
| Total | $4.414B | 100% |
These numbers produce perhaps the strongest finding in our study.
Roughly 68% of the 2026 capital we identified went into AI infrastructure, AI security, frontier research, and robotics.
In other words, most of the money did not go toward lightweight application companies.
It went toward companies building expensive technical foundations.
That includes data systems, cloud infrastructure, AI security, advanced models, robotics, and new approaches to machine intelligence.
This looks less like investors chasing a short-lived software trend and more like investors funding the construction of a new technology stack.
New York Is Becoming an AI Infrastructure City
The popular image of New York technology is still heavily tied to fintech, advertising, e-commerce, and enterprise software.
Those industries remain important.
But VAST Data, Modal, Pinecone, OpenRouter, and Dataiku show that New York increasingly has a serious AI infrastructure layer.
Together, the companies we classify primarily as AI infrastructure and enterprise AI platforms have raised more than $3 billion cumulatively.
VAST Data is the giant in the infrastructure group
VAST Data is one of the most important companies in the entire dataset.
The company announced a 2026 Series F at a $30 billion valuation and describes itself as building an AI operating system that unifies data infrastructure for large-scale AI systems. Its financing involved roughly $1 billion of combined primary and secondary capital.
Why is so much money available for something as seemingly unglamorous as data infrastructure?
Because powerful models are useless if companies cannot move, store, retrieve, govern, and serve enormous amounts of information quickly enough.
GPU capacity gets many of the headlines, but AI systems also create an enormous data problem.
VAST is effectively betting that companies will need a new data architecture built around AI workloads rather than trying to force those workloads into systems designed decades earlier.
Investors clearly believe that infrastructure layer can become extremely valuable.
Modal Shows Investors Want a New Cloud for AI
Modal offers another version of the same thesis.
The New York company raised $355 million in May 2026, giving it a reported $4.65 billion valuation. The company said its annualized revenue had moved above $300 million after growing fivefold since September.
Modal is not building another general-purpose cloud in the traditional sense.
Its pitch is that the infrastructure developers need for AI inference, reinforcement learning, model training, batch processing, agents, and large GPU workloads should work differently from ordinary cloud infrastructure.
That is an important distinction.
The first cloud boom was built around websites, databases, mobile applications, and software-as-a-service.
The AI cloud may be built around models that need GPUs for short bursts, huge parallel jobs, agents that spin up temporary environments, and workloads that change far more quickly.
Modal’s funding suggests investors believe a meaningful part of the cloud stack may be rebuilt around those needs.
Pinecone and OpenRouter Show the Infrastructure Stack Keeps Splitting
Pinecone has raised about $138 million and remains headquartered in New York City. Its vector database was designed to provide storage and retrieval infrastructure for AI applications.
OpenRouter sits at another layer.
Instead of storing the information that AI systems need, OpenRouter helps developers access and route requests across many different AI models. The company raised a $113 million Series B in May 2026, following a $40 million Series A, bringing total disclosed equity funding to about $153 million.
These companies demonstrate why simply asking which model will “win AI” misses a large part of the investment opportunity.
Companies are being created around every layer surrounding the models.
Someone must route them.
Someone must store their knowledge.
Someone must run their compute.
Someone must protect their data.
Someone must monitor what their agents are doing.
A large portion of New York’s AI funding is going into those supporting layers.
Cyera Shows That AI Security May Become Infrastructure Too
Cyera is the most heavily funded company in our dataset.
Its latest $600 million Series G valued the company at $12 billion. Together with a $400 million round earlier in 2026, Cyera raised roughly $1 billion in six months and has accumulated more than $2.3 billion overall.
The scale of that financing makes more sense when we look at what is happening inside large companies.
Enterprises are rapidly introducing copilots, agents, models, and automated workflows. Every one of those systems potentially touches corporate data.
That creates a difficult question: what should an AI agent be allowed to see?
Cyera argues that enterprises need a trust layer capable of understanding sensitive information and controlling how both humans and AI systems interact with it.
That moves data security closer to the center of the AI architecture.
Investors appear to agree.
In our strict 2026 dataset, AI security and governance alone account for about 23% of analyzed new capital, almost entirely because of Cyera.
This is one of the clearest signals for founders building for large businesses.
As enterprise AI gets more powerful, products that answer questions around permissions, auditability, data access, security, compliance, and agent behavior may become more valuable rather than less.
New York Is Suddenly Producing Frontier AI Labs
Perhaps the most surprising development in 2026 is the emergence of very young, very heavily funded research companies.
Flourish and General Intuition were both founded in 2025.
Together, they have already raised approximately $954 million.
That means two companies that barely existed a year ago represent more than 8% of all cumulative funding in our 21-company dataset.
More strikingly, their 2026 rounds represent about 18.6% of the strict 2026 equity capital we tracked.
Flourish raised $500 million before becoming a normal software company
Flourish emerged with one of the most unusual AI theses in New York.
The company is studying the human brain in an attempt to understand whether biological intelligence contains computational principles that could make artificial intelligence dramatically more energy-efficient and capable of continuous learning.
It reportedly raised $500 million at a $2.5 billion valuation, with backers including Jeff Bezos, GV, Lux Capital, and Catalio.
This is not ordinary software venture capital.
Investors are funding a scientific thesis.
That matters because it shows that New York investors and founders are now competing for something the city historically had less of: extremely expensive, long-horizon frontier AI research.
General Intuition is making a different bet on intelligence
General Intuition is approaching the problem from another direction.
Instead of learning primarily from text, its systems are being trained on actions and gameplay data.
The company raised $320 million in June 2026 at a $2.3 billion valuation, bringing its total disclosed funding to approximately $454 million.
By late August, reports suggested investors were already discussing another round at a far higher valuation. We did not include that possible round because it had not been confirmed as closed at our research cutoff.
This is another important methodological point.
Funding negotiations show investor appetite, but they are not the same as money that has actually reached the company.
Funding Velocity Shows How Fast the Market Has Changed
We created another metric to understand which businesses are receiving the sharpest recent acceleration.
We call it the Funding Freshness Ratio.
The calculation is simple:
Confirmed 2026 primary venture equity ÷ cumulative disclosed funding
A high ratio means a large part of the company’s total capital base arrived in 2026.
Chart: Companies with the highest 2026 funding freshness
| Company | Approx. cumulative funding | Strict 2026 equity used | Funding freshness |
|---|---|---|---|
| Flourish | $500M | $500M | 100% |
| Modal | $466M | $355M | 76% |
| Rogo | $311M | $235M | 76% |
| OpenRouter | $153M | $113M | 74% |
| Basis | $138M | $100M | 72% |
| General Intuition | $454M | $320M | 70% |
| Standard Bots | $287M | $200M | 70% |
| Profound | $155M | $96M | 62% |
| Taktile | $213M | $110M | 52% |
| Norm AI | $260M+ | $120M | ~46% |
| Cyera | $2.3B+ | $1.0B | ~43% |
This may be more useful than the raw funding ranking.
Older companies naturally have larger cumulative totals.
But Funding Freshness tells us where investors are dramatically increasing their exposure right now.
Modal, Rogo, OpenRouter, Basis, General Intuition, Standard Bots, and Profound stand out.
Those are very different businesses, yet they share one trait: investors believe their markets have entered an acceleration phase.
Wall Street Is Still One of New York AI’s Greatest Advantages
If infrastructure is one part of New York’s AI story, finance is the other.
AlphaSense, Rogo, Hebbia, Taktile, Basis, and Norm AI account for roughly $2.83 billion in cumulative funding in our professional-workflow category.
That number is particularly interesting because these companies do not all sell the same product.
Their common feature is that they attack information-heavy, high-value work.
AlphaSense has become the giant of financial intelligence
AlphaSense raised another $350 million in June 2026 at a $7.5 billion valuation.
The company also reported passing $600 million in annual recurring revenue during the first quarter of 2026.
Its cumulative funding is now about $1.747 billion.
AlphaSense illustrates one of the strongest types of AI business New York can produce.
The product is not merely generating text.
It sits on top of large amounts of valuable business information and helps professionals make expensive decisions.
That distinction is critical.
Companies may experiment casually with general AI products. They behave very differently when software becomes part of investment research, competitive intelligence, legal analysis, underwriting, accounting, or another decision where millions of dollars can be affected.
Those customers demand accuracy, security, integrations, audit trails, and reliable data.
Those requirements create barriers that are difficult for a simple AI wrapper to copy.
Rogo Is Turning Wall Street Itself Into a Distribution Advantage
Rogo is one of the strongest examples of a company that seems almost designed for New York.
The company builds AI specifically for investment banks, asset managers, and other financial institutions.
In January 2026, Rogo announced a $75 million Series C. Just three months later, it announced another $160 million Series D, bringing publicly reported total funding above $300 million.
That is an extraordinary fundraising pace.
It also tells us something about the value of specialization.
General-purpose models already know a great deal about finance.
That has not made Rogo unnecessary.
Instead, customers appear willing to pay for systems that understand their workflows, connect with institutional data, fit security requirements, and perform the exact work financial professionals need.
This is one of the biggest lessons in our dataset.
Having access to powerful models does not eliminate vertical AI.
It can make vertical AI more valuable because specialized companies can build deeper products on top of increasingly capable models.
Hebbia Shows the Opportunity—and the Risk
Hebbia has raised about $161 million, including a $130 million Series B announced in 2024. Its Matrix product became known for helping investment and professional-services teams analyze large sets of documents.
But Hebbia also demonstrates why large funding rounds do not guarantee permanent leadership.
By 2026, competing AI products had adopted similar document-analysis and workflow features, pushing Hebbia to launch a substantially redesigned version of Matrix.
That competitive pressure contains a useful warning.
Features spread quickly in AI.
A startup cannot assume that one impressive workflow will remain unique.
The stronger defenses are proprietary data access, difficult integrations, customer relationships, workflow depth, trust, distribution, and the ability to keep improving faster than competitors.
Taktile Shows Decision-Making Is Becoming an AI Category
Taktile raised $110 million in June 2026.
The company focuses on helping financial institutions automate decisions around areas such as lending, claims, customer onboarding, and financial crime.
Its total funding has reached approximately $213 million.
This category may become much larger than it sounds.
Many industries are essentially enormous collections of decisions.
Should this loan be approved?
Should this claim be paid?
Does this transaction look suspicious?
Should this customer be accepted?
Should a case be sent to a human?
As models improve, the market may move from AI that simply provides information to AI that participates directly in those decisions.
The companies that manage that transition safely could become extremely important enterprise platforms.
Norm AI Shows Why Compliance Is Becoming a Software Market
Norm AI raised $120 million in 2026 at a $1.2 billion valuation, taking total funding above $260 million.
Its core idea is unusually well suited to New York.
Instead of treating regulations as documents employees must manually read and interpret, Norm attempts to embed legal and regulatory rules into AI systems.
The company has also developed an affiliated AI-native law firm.
That approach sits at the intersection of several New York strengths: finance, law, regulated industries, and enterprise technology.
Compliance has historically been expensive because regulations must be interpreted and applied to a huge number of individual actions.
AI can potentially reduce that cost.
But it also creates new regulatory risk.
That creates a powerful two-sided opportunity for companies that can make AI both more capable and more compliant.
Basis Shows Why “Boring” Industries Can Produce Huge AI Companies
Accounting may not sound like the most exciting area of artificial intelligence.
Investors clearly disagree.
Basis raised $100 million at a $1.15 billion valuation in February 2026, bringing total funding to roughly $138 million.
Its agents are designed to handle accounting, audit, and tax workflows.
That is exactly the kind of market that often looks unattractive until someone understands the economics.
Accounting work is repetitive enough to automate, complicated enough to require expertise, and important enough that customers will pay for reliability.
There is also an enormous installed base of firms already spending large amounts of money on labor.
That creates a straightforward economic argument for AI.
If software can safely perform a meaningful percentage of work currently requiring expensive professional time, customers do not need to be persuaded that productivity has value.
They can calculate it.
Runway Keeps New York at the Center of Generative Media
New York’s media industry gives the city another natural AI cluster.
Runway remains the clearest example.
The company raised $315 million in February 2026 at a reported $5.3 billion valuation, taking cumulative funding to approximately $860 million.
Runway began with AI creative tools but has steadily pushed deeper into model research.
Its ambitions now extend toward world models—systems intended to understand and simulate environments rather than simply generate attractive video clips.
That matters because generative video is extraordinarily expensive to build.
It requires model research, enormous amounts of compute, specialized data, and constant improvements as competitors release new systems.
Runway’s funding level reflects those capital requirements.
It also demonstrates why New York may remain unusually competitive in creative AI.
The city already has advertisers, publishers, filmmakers, agencies, brands, designers, creators, and media companies.
A generative-media company in New York is not building thousands of miles from its potential customers.
It is often building in the same city.
Mirage Shows There Is Room Below the Frontier-Model Layer
Mirage, formerly known as Captions, occupies a different position.
Its products make AI video creation easier for creators, marketers, and businesses.
In March 2026, the company announced $75 million in growth financing, taking reported total funding above $175 million. It said more than 20 million users had used Captions and that more than 250 million videos had been created using the platform.
Runway and Mirage therefore represent two related but distinct opportunities.
One is pushing heavily into foundational generative-video research.
The other is packaging AI video capabilities around practical creation and business workflows.
Both can succeed.
The important question for startups is where they can create an advantage that survives as underlying models improve.
AI Is Moving Into the Physical World
Standard Bots adds another important category to New York’s AI map.
The company raised a $200 million Series C at a $1 billion valuation in June 2026 to expand its AI-native industrial robotics business.
Public funding databases place cumulative financing at roughly $287 million.
Robotics is fundamentally different from software.
A software agent can fail and be restarted.
A robot operates in a physical environment where mistakes can damage products, equipment, or people.
That raises the technical bar.
It also raises the potential value of companies that solve the problem.
The inclusion of Standard Bots and General Intuition among the most aggressively funded companies in our dataset suggests investors expect the next major AI frontier to move beyond screens.
EliseAI Demonstrates the Power of Vertical AI
EliseAI may be one of the most important companies for understanding New York’s vertical AI strategy.
The company initially focused on automating communication and operations for property managers before expanding into healthcare.
It has raised more than $400 million in confirmed funding and reported passing $200 million in annual recurring revenue in 2026.
Reports in August said investors were discussing another $300 million round at a $3.7 billion valuation.
We have not counted that amount because the transaction remained unconfirmed at our cutoff.
Even without the potential new round, EliseAI illustrates why vertical AI can become large.
Real estate management is filled with repetitive communication: leasing questions, tour scheduling, maintenance requests, billing issues, resident communication, and administrative work.
Healthcare administration has many similar patterns.
These are huge industries with expensive labor and fragmented workflows.
An AI company does not need to replace the entire industry.
It only needs to automate a meaningful percentage of a large recurring workload.
Cadence Shows Investors Also Want AI Tied to Measurable Outcomes
Cadence raised $100 million in June, bringing total funding to roughly $241 million.
The company works with health systems to manage chronic care remotely and uses technology and AI to identify patients who need attention.
Healthcare can be difficult for software companies because simply making work faster is often not enough.
Buyers want better clinical or financial outcomes.
That requirement creates a higher barrier to entry, but it can also produce a much stronger business once the technology demonstrates value.
The same logic applies across many regulated New York verticals.
The more serious the consequence of a decision, the less likely customers are to accept a shallow AI product.
Profound Is Betting That AI Changes How Customers Discover Brands
Profound may look smaller next to billion-dollar infrastructure companies, but its growth rate makes it one of the most interesting companies in the dataset.
The startup raised a $96 million Series C at a $1 billion valuation in February 2026, bringing total funding above $155 million less than two years after launch.
Its market exists because AI answer engines are changing search behavior.
Brands have spent decades learning how to appear in Google.
They now need to understand how they appear when people ask ChatGPT-style systems for recommendations, comparisons, and answers.
That creates an entirely new marketing problem.
Profound is betting that “visibility inside AI answers” becomes a category as important as search-engine visibility became during the Google era.
Its funding suggests investors believe that possibility is becoming real.
One of the Strongest Findings: NYC AI Is Mostly Enterprise AI
We manually classified the companies in our dataset based on their main customer and product model.
At least 17 of the 21 companies primarily sell infrastructure, enterprise software, institutional tools, or regulated-industry technology.
Those 17 businesses account for approximately 83% of the cumulative capital in the dataset.
NYC Tech Journal enterprise-orientation analysis
| Metric | Result |
|---|---|
| Companies in dataset | 21 |
| Primarily enterprise/infrastructure companies | 17 |
| Share of companies | 81% |
| Approx. share of cumulative funding | 83% |
This may be the most New York-specific result in the entire analysis.
The city’s AI boom is overwhelmingly tied to business problems.
That fits New York.
A startup founder here has immediate access to some of the world’s largest financial institutions, law firms, media companies, healthcare systems, advertisers, insurers, real estate operators, accounting firms, and corporate headquarters.
Those businesses have large budgets.
They also have complicated problems.
That combination is ideal for vertical AI.
Younger NYC AI Companies Are Capturing Capital Extremely Quickly
Age provides another useful lens.
Companies founded in 2023 or later account for only around 14% of cumulative funding in our dataset.
But they captured roughly 28% of the strict 2026 equity capital we identified.
The shift becomes even clearer when we look only at companies founded in 2025.
Flourish and General Intuition together account for approximately $820 million of 2026 equity financing, or nearly 19% of the analyzed 2026 total.
Chart: New companies are gaining funding share faster than cumulative share
| Company age group | Share of cumulative dataset funding | Share of analyzed 2026 capital |
|---|---|---|
| Founded 2023 or later | ~14% | ~28% |
| Founded before 2023 | ~86% | ~72% |
This suggests investors are not simply continuing to finance companies that became large during the 2010s and early 2020s.
They are rapidly creating a new generation of potential market leaders.
That is healthy for the ecosystem.
A city cannot become a durable technology center if all of its large companies were founded years ago.
It needs new companies constantly entering the top tier.
In 2026, New York appears to be doing exactly that.
The Wider NYC Funding Market Supports the Same Conclusion
Our company-level dataset is not the only evidence that AI is taking a larger share of New York venture capital.
AlleyWatch reported that AI companies received about 46.6% of all NYC capital in February 2026.
Its April analysis put the share at roughly 66.7%.
In May, AI companies received roughly $1.32 billion, representing about two-thirds of the capital tracked that month.
Then June produced an extraordinary $2.61 billion across 41 AI-focused companies, approximately 55.5% of all NYC funding that month.
So the trend does not depend on one company or one giant round.
AI is repeatedly taking somewhere around half or more of the city’s venture dollars during major funding months.
Why New York’s AI Boom Looks Different From Silicon Valley’s
New York does not need to become a copy of San Francisco to become one of the world’s most important AI cities.
Its strongest strategy may be the opposite.
The city can build around advantages Silicon Valley cannot easily recreate.
Wall Street is here.
Major advertising agencies are here.
A huge portion of the media industry is here.
Some of the world’s most powerful law firms are here.
Real estate is one of the city’s largest industries.
New York has major hospital systems, insurers, consulting firms, corporate headquarters, fashion companies, accounting firms, and institutional investors.
Those organizations provide something every AI startup eventually needs: customers with difficult problems and money to spend.
That may explain why so many of New York’s best-funded AI businesses are not trying to build another general consumer assistant.
They are automating expensive work.
What Founders Can Learn From Where the Money Is Going
The funding data creates several practical lessons for anyone building an AI company in New York.
Start with an expensive problem, not an impressive demo
The strongest vertical companies in our dataset attack work that already costs customers serious money.
Rogo targets financial professionals.
Basis targets accountants.
Norm targets legal and compliance work.
EliseAI targets property and healthcare operations.
Taktile targets financial decision-making.
Investors can understand the economic argument because customers already have a budget attached to the problem.
A technically impressive AI product without a clear economic buyer is much harder to finance.
Go deeper into the workflow
One prompt box is easy to copy.
A deeply integrated workflow is harder.
The best vertical AI businesses increasingly connect to company data, existing software systems, permissions, rules, approvals, and downstream actions.
That changes the product from “AI that gives me an answer” to “AI that helps complete the job.”
The latter is much more valuable.
Treat reliability as part of the product
In serious industries, accuracy is not simply an engineering metric.
It is part of the buying decision.
A lawyer, accountant, banker, insurer, or healthcare provider cannot use a system that works impressively most of the time but unpredictably fails on important cases.
Products designed for these markets need auditability, human review, testing, source visibility, and controls.
That may sound less exciting than model performance.
It is often what converts a demonstration into a large enterprise contract.
Use New York as a product-development advantage
Being close to customers matters most when the market is complicated.
A founder building for investment banks can learn faster by speaking to bankers every week.
A legal AI company benefits from constant interaction with attorneys.
A property-management AI company can understand real workflows by working directly with large building operators.
New York gives founders access to exactly those users.
That proximity should influence where companies choose to build, hire, and sell.
What Investors Appear to Be Rewarding in 2026
Our dataset suggests that venture firms are paying up for four characteristics.
The first is technical depth.
VAST Data, Modal, Cyera, Flourish, General Intuition, Runway, Pinecone, and OpenRouter are not businesses that can be rebuilt over a weekend with access to an API.
The second is high-value distribution.
AlphaSense, Rogo, EliseAI, Taktile, Norm, Basis, and Hebbia sell into customers where one strong relationship can become a very large account.
The third is proprietary workflow knowledge.
A general model knows what accounting is.
Basis needs to know how accounting work actually gets completed.
A general model understands finance.
Rogo has to understand how analysts and bankers perform their jobs inside major institutions.
That difference is where much of the product value exists.
The fourth characteristic is market size.
Investors appear willing to fund narrow-looking products when the underlying industry is enormous.
Accounting is specialized.
Property management is specialized.
Compliance is specialized.
Investment research is specialized.
But each represents billions of dollars in annual labor and software spending.
A vertical can be narrow and still support a giant company.
What We Would Watch During the Rest of 2026
Several unresolved questions could change the ranking quickly.
EliseAI is the clearest immediate example.
A confirmed $300 million round would materially increase both its cumulative funding total and the share of 2026 capital flowing into vertical operational AI. Current reporting still describes the transaction as being discussed, so we have left it outside our calculations.
General Intuition is another company to watch.
Reports in late August said investors were discussing another financing at a roughly $6 billion pre-money valuation only weeks after the startup closed its $320 million Series A.
If that deal closes, the amount of capital flowing into New York frontier AI could rise again.
The bigger question, however, is whether the Series A and seed market begins producing more companies capable of joining this group.
Large late-stage rounds make headlines.
Ecosystem depth is built much earlier.
The Biggest Risk: Funding Can Hide Weak Businesses
Large AI funding rounds are easy to interpret as proof that a company has won.
They are not.
Capital gives companies more time, more computing power, better recruiting options, larger sales teams, and greater ability to expand.
It does not guarantee durable demand.
Some AI categories are changing so quickly that products can become outdated between funding rounds.
Underlying models are improving.
Open-source alternatives are improving.
Enterprise buyers are learning which AI tools they actually need.
Large software vendors are rapidly adding competing functionality.
The result is an unusually unforgiving environment.
The companies that justify these valuations will need more than strong fundraising.
They will need revenue growth, high retention, measurable customer value, improving unit economics, and products that remain differentiated as models become cheaper and more powerful.
What New York’s Funding Map Says About the Next AI Cycle
The most useful conclusion from our dataset is not that New York has raised billions of dollars for AI.
It is that the money forms a recognizable pattern.
Investors are financing infrastructure because AI requires a new technical foundation.
They are financing security because powerful agents create new risks.
They are financing frontier labs because the industry is still searching for approaches beyond today’s large language models.
They are financing finance, legal, accounting, real estate, and healthcare companies because those industries contain expensive workflows that AI can automate.
And they are financing generative media because New York already has one of the deepest creative economies in the world.
These themes reinforce one another.
Infrastructure companies make application companies possible.
Application companies create demand for infrastructure.
Enterprise AI creates security problems.
Security products make enterprises more comfortable deploying AI.
Better models make vertical agents more capable.
Vertical customers generate more revenue that can fund deeper product development.
That is what a technology ecosystem begins to look like when it moves beyond experimentation.
Final Takeaway: NYC Is Building an AI Economy Around Real Work
Our original analysis found at least $11.58 billion in cumulative disclosed funding across 21 of New York’s most heavily funded private AI companies.
The top three account for roughly 47% of that capital.
The top 10 account for about 81%.
More than $4.4 billion of conservatively counted primary venture equity flowed into this small group during 2026.
Roughly 68% of that 2026 money went toward infrastructure, security, frontier AI, and robotics.
And roughly 83% of the cumulative capital in our dataset belongs to companies that primarily serve enterprises, infrastructure markets, or institutional customers.
Those numbers tell a consistent story.
New York is not winning AI by building endless copies of consumer chatbots.
It is building around the things New York already does unusually well: money, media, law, property, healthcare, enterprise operations, complex information, and high-value decisions.
At the same time, companies such as VAST Data, Modal, Flourish, General Intuition, Pinecone, OpenRouter, Runway, Cyera, and Standard Bots are widening the city’s ambitions into infrastructure, frontier research, and physical AI.
That combination may be New York’s real advantage.
The city does not have to choose between deep AI technology and practical business applications.
In 2026, venture capital is increasingly funding both.
