A striking trend in US tech circles has been the growing number of top researchers choosing to leave the tech giants to start their own artificial intelligence companies. The trend, analysts say, is driven both by management reasons within the tech giants and by investors’ willingness to bet on start-ups.
‘a fleeting profit window’
Top researchers at big tech companies such as Meta, Google and OpenAI are leaving to start their own AI companies. One important reason, the report says, is that investors have high hopes for the commercial potential of these early AI labs. Many start-ups have raised hundreds of millions of dollars in their first few months.
File Photo: OpenAI logo.
On April 27, Silver, a former Google Deepmind researcher, announced that he had raised $1.1 billion in seed rounds for his months-old startup, Ineffable Intelligence. In the past year, former employees of OpenAI, DeepMind, Anthropic and Xai have raised hundreds of millions of dollars for months-old ventures.
Orize, a French venture capital firm, has invested in AMI Labs, founded by Yang Likun, a Turing prize-winning former chief AI scientist at Meta. Oerizer executives say the competition among the big A.I. companies has led to opportunities for smaller, more nimble companies. The executive added that when AI companies are in competition, they tend to narrow their focus, creating a“Vacuum” in research and development, new architectures, agents, and vertical domain-specific models are all being shelved. It is not that these areas are unimportant, but that they do not help companies to get ahead in the short-term race.
According to Business Insider, entrepreneurship has become a way for employees at some technology companies to chase opportunities. Especially in AI, rapid progress is opening the door to new business ventures.
The wave of departures is also linked to the internal management of tech giants. Last year, for example, Nicole Ferragonio and Joe Lukes decided to leave Amazon to start an artificial intelligence company together. The pair had been planning for months, but it was Amazon’s elimination of flextime in January 2025 that really set Ferragonio’s heart. Lukes believes that the rapid rise of“Game-changing” artificial intelligence technology, this is a fleeting profit window period.
Why investors like AI start-ups
AI labs, founded by top researchers, is attracting a rush of investors. In 2026, venture capital firms invested $18.8 bn in AI start-ups established since early 2025, according to Dealroom, an investment and financing data platform. At that rate, that figure could surpass the $27.9 billion in financing for start-ups in 2024.
Executives at Orizer say top talent who have worked in the tech giant’s cutting-edge labs have“Unique” insights. “They know what works at scale and they know exactly what is missing internally. That’s where the opportunity lies.”
AI labs run by big companies are trying to justify astronomical valuations, says a partner at German venture Capital firm HV Capital, their focus on commercial goals limits the freedom of top researchers.
Liang Huaixin, a researcher at the national security and Governance Research Institute of the Beijing University of International Business and Economics, told a reporter from the Global Times on the 29th that, on the one hand, AI start-ups can realize the in-depth development of a certain aspect of AI technology. If the relevant talents are in a large AI factory in the United States, it is difficult to obtain sufficient resource support, this specialization development path is one of the important reasons why the society is optimistic about AI start-ups. At the same time, AI start-ups can also combine AI technologies that are already relatively mature with specific social application scenarios. On the other hand, compared with the large AI factories in the United States, AI start-ups have relatively controllable innovation costs, and can eliminate the limitations of the company’s management level to achieve the greatest degree of incremental development. Therefore, at this stage, the enthusiasm of the Society for AI start-ups is relatively high.
Some senior researchers set up start-ups in order to explore the“Gaps” in the AI industry. Recursive Intelligence was founded in September and closed two rounds of funding in December and January for a total of $335 million. Its core business is developing AI tools for chip design, CNBC reported. The founders were involved in DeepMind’s AlphaChip project, which aims to automate the entire chip design process.
According to founder Anna Goldie, start-ups are more likely to be seen by customers as partners rather than competitors than as industry giants. “The chip companies are willing to hand over their core intellectual property to us as long as we remain neutral. If we were at Google, that would not be possible.”
Huge investment raises concerns
The departure of senior researchers to start businesses has also been accompanied by concerns about the ability of us tech giants to make a return on their huge investments. According to the Wall Street Journal, Openai recently missed its targets for new users and revenue, a setback that has raised concerns among the company’s leadership, they worry about whether the company can afford to spend heavily on data centers. People familiar with the matter said Ms Fryer, OpenAI’s chief financial officer, had told other company leaders she was concerned that the company might not be able to pay for future computing contracts if revenues did not grow fast enough.
Microsoft, Alphabet, Amazon and Meta, the four biggest tech companies, plan to invest more than $650 billion in AI by 2026, according to new estimates. However, the huge spending plans have put pressure on capital markets. “The scale of Capex is mind-boggling,” says Tilney, head of funds at intercom asset management, even as their combined annual revenues rise 14% to $1.6 trillion, that would be hard to sustain.