Meta, Google, and OpenAI’s top staff are leaving to launch AI startups
Synopsis
Senior AI researchers are leaving Big Tech in growing numbers to launch their own startups, and raising record amounts fast. Former Google DeepMind researcher David Silver raised $1.1 billion for Ineffable Intelligence. Meta's former AI chief Yann LeCun raised $1 billion for AMI Labs. VCs have already put $18.8 billion into AI startups founded since the start of 2025. Researchers say commercial pressure at the big labs is squeezing out serious long-term research, and investors are backing the people willing to do it on their own terms.
Tech is losing the biggest names in AI research as they move straight into billion-dollar funding rounds at Big Tech. This race for AI dominance is opening up new space for smaller, faster startups that are rushing to fill, and investors aren’t able to write the cheques fast enough.
Key Highlights
- David Silver, the ex-Google DeepMind researcher recently raised a gargantuan $1.1bn seed round for his weeks-old startup.
- In another fallout, AMI Labs had raised $1 billion in March, months after Meta’s chief AI scientist Yann LeCun stepped down as founder.
- Investors have invested $18.8 billion in startups founded since the dawn of 2025, heading for a tally that could be bigger than all of last year's $27.9 billion into AI startups overall.
- In months, ex-Anthropic and DeepMind staff scored $335 million for chip design AI startup Ricursive Intelligence.
The Builders Behind Big Tech AI Are Now Building Their Own
Something significant has been happening in the AI industry recently. The most battle-hardened researchers, the people shaping the core of what Google, Meta, OpenAI and Anthropic are, are quitting. And the money coming after them is like no one has seen at this level of play before. Ex-Google DeepMind researcher David Silver raked in a record $1.1 billion for his new company Ineffable Intelligence, months after quitting Meta’s AI chief for years, Yann LeCun left to establish AMI Labs, which has raised $1 billion in March. Also from DeepMind is Tim Rocktäschel, hanging out his new shingle and raising to $1bn for Recursive Superintelligence.
Why They Are Leaving
The large AI labs feel intense pressure to ship things quickly and meet revenue goals. That does not leave much space for research that fails to yield results quickly. It’s a case of “You narrow focus when you were racing,” said Elise Stern, a partner at venture capital firm Eurazeo, who invested in AMI Labs. “That leaves a gap: entire fields of research are getting pushed to the side, not because they’re unimportant but simply because they fall behind in the race right now, and banks don’t like that.”
Now, as researchers seek to tackle more difficult, extended projects find that the only option is to leave and start anew. Venture capitalists have already invested $18.8 billion into new AI companies just this year, on pace to surpass last year's total of $27.9 billion.
What This Means for the AI Industry
These companies are not just replicating what they were doing at their previous employers, they are tackling problems left behind by the large labs. Last summer, two ex-Anthropic and DeepMind staff launched Ricursive Intelligence, which makes AI for chip design and scored $335 million in funding only a few months later.
Founded in October 2023 by former employees of Anthropic and xAI, Humans& finished its first funding round at $480 million mere weeks after launching. As founder Anna Goldie puts it, “To earn the right to work with chipmakers on their most valuable IP, we must be Switzerland, and that would be impossible if our funding came from Google.” It does not get better with longer-term money alone for the big labs; losing people of this stature is trivial and yet a persistent problem.
FAQs
- What was the reason why top AI researchers left Big Tech?
In the shrinking space of exploratory research under commercial pressure, startups provide the autonomy and funding to explore what is being overlooked by entrenched labs.
- How Much Are These Start-Ups Raising?
Hundreds of millions to over a billion, often within months after launch at C stage.
- Are they going up against Google and OpenAI directly?
Mostly not, they are going after very specific openings the big labs have left behind; like chip design AI, autonomous research, and physical world AI.
- What is reinforcement learning?
AI that learns from experience instead of human data, the pitch behind many of these start-ups’ bets is that it trumps a large language model approach on demonstrable tasks.
Follow Inspirepreneur Magazine for daily global business news.
At Inspirepreneurs Magazine, covering entrepreneurship, business failures, and the human stories behind the world's most ambitious founders. She writes at the intersection of strategy and storytelling.
You Might Also Like
EQT Partners with Google Cloud to Expand AI Across 300 Portfolio Companies
Grocery Prices at Woolworths Are Going Up, New Survey Reveals
