AI startups are making their home in New York. Can they turn it into an AI powerhouse?
New class of startups setting up shop in NYC, hoping to tap its concentration of businesses ready to buy AI apps
Artificial intelligence is transforming the technology sector in New York City—into a booming one worth billions and rivaling San Francisco in innovation, investment and talent.
The Big Apple is making a bid to become a premier hub for generative AI, according to investors, founders and businesses with ties to the region. That’s because of a combination of public and venture-capital funding, tech worker migration and, most important, a critical mass of companies eager to buy the technology.
New York City-based generative AI companies raised $2.7 billion since the start of 2021, according to Pitchbook Data. So far this year, investors have put $200 million through 27 deals into generative AI startups in New York City, compared with $400 million through 35 deals in the category in San Francisco, the data provider said.
Funding for all AI firms in New York City was $26.8 billion from 2021 to the first quarter of 2024, compared with $54 billion in San Francisco, Pitchbook said. San Francisco and New York are in first and second place among all cities worldwide for AI investment, followed by Beijing and London, according to the data provider.
In January, New York state announced plans to build up its AI research bona fides through a $400 million public-private partnership dubbed “Empire AI.” Data center operator Databank unveiled a new facility in Orangeburg, N.Y., last week with Coreweave, a New Jersey-based cloud startup valued at $19 billion , as its anchor tenant.
Since the generative AI boom kicked off in late 2022 , businesses in New York have become some of the first paying customers of large-language models and tools such as Openai’s CHATGPT and Microsoft ’s Copilot assistant, even as they remain errorprone and expensive to use.
Customers, customers, customers
Whereas prior technology booms such as social media and mobile set off innovation in the consumer market, tech leaders say generative AI favors businesses because it is ideally suited for transforming large organizations. Generative AI can create various types of content, from text to video, because of the vast amounts of data it has been trained on—allowing developers to build applications like chatbots and design tools that can boost worker productivity.
Industries heavily present in New York, including financial services, healthcare and law, have been among the first to adopt generative AI.
Because those sectors are regulated, and therefore have a good handle on their internal data, they can more quickly apply large-language models to their operations, said Matt Wood, vice president of AI products at cloud business Amazon Web Services.
AI startups have flocked to New York to tap that customer base, and be hands-on in helping chief information officers use the technology, said Grace Isford, a partner at Nyc-based venture firm Lux Capital.
Andrew Hoh, co-founder and chief product officer of Lastmile AI, rents office space in the Empire State Building for his nine-person startup, which makes a platform to help developers build generative AI applications. Founded in 2023, Lastmile AI uses the space as a hub for meeting financial services customers, who are often just a few blocks away.
“It’s a little bit easier when you’re able to shake someone’s hand in-person, especially for a traditional business,” Hoh said. “That’s the strategic advantage that we had.”
Many AI startups in New York aren’t developing massive models like San Francisco’s Openai is. Instead, they’re selling ready-to-use AI tools for sectors like banking and healthcare, said New York-based Insight Partners Managing Director Lonne Jaffe .
Cristóbal Valenzuela, co-founder and chief executive of Runway, said the startup’s media and advertising customers are primarily in New York, where he and his co-founders met. Runway, which has raised $237 million for its generative AI video-editing and -creation tools, moved from Tribeca to a larger office in Union Square.
“New York is actually better suited to drive a lot of innovations for the next couple of years,” Valenzuela said. “You’re going to move through stages of application use cases that require people to have a deeper understanding of the industry they’re working on, and in San Francisco, it’s just too much tech.”
Talent and research
New York gained more relocated tech workers than any other U.S. city in 2023, according to an analysis of Linkedin data by venture firm Signalfire. But San Francisco continues to draw in AI talent at a faster rate. AI engineers based out of New York accounted for about 15% of all AI engineers working at startups last year, second to San Francisco, where more than 41% were located, the Signalfire report said.
Edo Liberty, founder and CEO of Pinecone, said New York has for years been underappreciated as an AI talent hub. The five-year-old firm, which builds database technology that helps businesses use their private data with large-language models, occasionally opens its Midtown office to younger startups in need of a corporate crash pad.
“I can’t say how many times a month an AI startup asks me if they can hang out in our office for a week or two until they find a space,” Liberty said.
The AI research labs of public tech companies , including Meta Platforms, Deepmind Technologies, a subsidiary of Google parent Alphabet , and Palantir Technologies , have been key sources and anchors of talent, along with local academic institutions such as Cornell Tech, Columbia University and New York University, where Meta chief AI scientist Yann Lecun is a professor.
New York is also a top destination for AI talent from Europe, Israel and Asia, who prefer to live in the city over San Francisco and other emerging tech hubs , some founders say. Runway’s Valenzuela, who moved from his hometown in Chile, said the company has sponsored work visas for about 20 researchers and engineers—roughly a quarter of the employees—all of whom want to live in New York.
Will AI ‘save New
York’?
Development of the generative AI sector couldn’t come soon enough for New York, said Dylan Reid, a New York-based partner at Zetta Venture Partners, a venture firm based in San Francisco.
“I think AI will save New York way more than San Francisco,” he said. “The New York startup ecosystem was really hurting.”
New York had been home to numerous high-profile startups in e-commerce, media, crypto and fintech that underperformed , Reid said. In recent years, the successful public offerings of database-software companies Mongodb in 2017 and Datadog in 2019 , have been cited as proof for entrepreneurs to take a leap of faith in New York.
“In 2014, a lot of people doubted that we could build a big, successful tech company in New York,” said Dev Ittycheria , Mongodb ’s CEO.
“But I’ve always found that access to both high-quality customers and investors is second to none in New York.”
San Francisco—home of Openai and many of its researchers—remains the de facto generative AI watering hole and startup capital. But new AI labs and venture- capital offices in the Big Apple are drawing people in. Over 150 AI firms are based or have satellite offices in New York, according to Lux Capital.
“I expect we’ll see a ton of really exciting AI startups out of New York,” Reid said. “Right now, it’s more talentheavy and less startup-heavy.”
Hugging Face, developer of the popular open-source AI model-sharing platform, opened up shop in Brooklyn’s Dumbo neighborhood—a decision co-founder and CEO Clem Delangue described as “you don’t need to follow the Silicon Valley hype and move there to succeed.”
Martin Kon, chief operating officer of Openai rival Cohere, said the company’s recent decision to open a New York office was a no-brainer. “We need to be where our customers, top AI talent, partners and investors are, and New York is a top priority,” Kon said.
Asense of disquiet is sweeping over the old continent. Innovative, low-cost Chinese producers are taking on Europe’s venerable industries and unsettling its policymakers. Europe has played only a bit-part in the tech revolution: the market value of America’s “magnificent seven” tech giants is about the same as that of the combined stockmarket capitalisation of the eu’s 27 members. In an interview with us last week, President Emmanuel Macron offered his own diagnosis. There can be no great power without economic prosperity and technological sovereignty, but “Europe does not produce enough wealth per capita.” It must become an attractive place to invest and innovate. This requires vast amounts of capital—and a well-oiled financial system that channels savings to promising investment opportunities across the continent.
The trouble is that European finance remains inefficient and bound by national borders. Pressing ahead with banking and capital-market reforms is thus more important than ever.
A decade ago European banking was on its knees. The sovereign-debt crisis in the south exposed an infernal doom loop. Because banks held a lot of sovereign debt and governments had to bail out banks in difficulty, trouble at one infected the other. Lenders were unprofitable, unloved by investors and saddled with non-performing loans. Today those bad loans have been shed and profits have recovered. The share price of Unicredit, one of Italy’s largest lenders, has outperformed that of Meta this year. Big banks are now subject to European supervision and regulation, rather than a patchwork of national measures. Yet banking on the continent remains cumbersome and parochial.
Europe’s banking union, first proposed in 2012, remains incomplete, mainly because a common deposit-insurance scheme has yet to be set up. One result is that the doom loop retains its power. Another is that too little cross-border activity and consolidation takes place. Regulators fear that if a bank collapses, they will be on the hook for loans made to dodgy borrowers beyond their borders. Without a common deposit-insurance scheme, governments require banks to use ring-fencing regimes to hoard liquidity that could have been more profitably deployed elsewhere. The time to press on with such a scheme is surely now, when non-performing loans are low everywhere and even southern banks are in good health.
Banks are not the only source of finance. More must also be done to create a European capital market, which can help spread risks. National markets are underdeveloped: according to the imf, only 30% of companies’ financing comprises tradable securities in the euro area, compared with two-thirds in America. As a consequence, Europeans’ vast savings are locked up in bank deposits, small firms struggle to obtain finance and entrepreneurs bear too much risk. By one estimate, a shock of one percentage point to national GDP growth results in a 0.8-point fall in consumption in the eu, but only a 0.18-point fall in America, where the pain is shared by investors, creditors and the government. No wonder Americans are more go-getting.
Although a proposal for a European capital-markets union was put forward in 2015, not much has been achieved. That is because regulations touching things like insolvencies, tax and disclosure have to be harmonised for an investor to see no difference between a security in Greece and one in Germany. Compared with whacking tariffs on Chinese cars, this is neither eye-catching nor easy. It is the spinach of public policy.
Eat your greens
Tariffs cheat European consumers by raising the cost of otherwise good and cheap products. By comparison, financial reforms channel Europe’s large pot of savings into profitable investments. They should encourage innovation and make the economy more resilient, by spreading risk more widely. Europe rightly worries that it is falling behind as America and China charge ahead in the race for technological supremacy. But, as with the two superpowers, one of its main strengths is size. To keep up, Europe should make the most of it.
Many AI startups
in New York aren’t developing massive models. Instead, they are selling ready-touse AI tools