2 min read AI-generated

Big Tech Researchers Are Founding AI Startups Left and Right — $18.8 Billion in One Year

Copy article as Markdown

Top AI minds are leaving OpenAI, DeepMind, and Meta to start their own labs. Investors are throwing billions at them. What this means for the industry.

Featured image for "Big Tech Researchers Are Founding AI Startups Left and Right — $18.8 Billion in One Year"

There’s a new trend in the AI industry, and it’s measured in billions: the brightest minds at OpenAI, DeepMind, Anthropic, and xAI are quitting — and weeks later founding their own companies that immediately attract absurd amounts of funding.

The numbers

According to CNBC, investors have already poured $18.8 billion into AI startups founded since early 2025 — just in 2026 alone. That’s on pace to surpass the $27.9 billion that flowed into comparable startups throughout all of 2025.

The most prominent recent examples: David Silver, the mind behind AlphaZero at DeepMind, raised a record $1.1 billion seed round for his months-old startup Ineffable Intelligence. Tim Rocktäschel, another former DeepMind researcher, is reportedly raising up to $1 billion for Recursive Superintelligence.

At OpenAI, Bill Peebles and Kevin Weil recently departed — Peebles led Sora, Weil was Vice President for Science. Both are considered likely founders of their own labs.

Why they’re leaving

At its core, the issue is straightforward: the big AI labs are increasingly focused on commercial goals. OpenAI’s valuation north of $850 billion, Anthropic’s revenue growth from $9 billion to $30 billion ARR in a single year, Meta’s multi-billion infrastructure plans — all of that creates pressure that doesn’t sit well with open-ended research.

At the same time, it’s never been easier for an AI researcher to start a company. Investors are lining up. If you have a strong track record at one of the major labs, your term sheet is practically signed before your company even has a website.

What it means

For the established labs, the talent drain is a real problem. Models are getting more complex, teams need to be more specialized — and those exact specialists are the ones walking out the door.

For the industry as a whole, though, it could be a net positive. More independent labs mean more approaches, more experiments, more diversity in research. Not everyone will succeed, but competition drives innovation. And perhaps these startups will produce the next big breakthroughs — beyond the same scaling playbook the giants keep running.


Sources: CNBC — Big Tech staff leaving for AI startups, TechCrunch — Musk relitigates an old friendship