Why Top AI Leaders Are Exiting Big Tech to Build Their Own

News

1. Nexus Closes Eighth Fund at $700 Million for Early-Stage Bets in India and US

2. Amazon to Pump $12.7 Billion into India’s AI and Cloud Expansion

3. Anthropic Debuts Interviewer Tool to Understand How People Really Use AI

4. IPL Tops Google’s Search List for 2025 as AI Breaks Into India’s Top 10

5. TeamLease Names Suparna Mitra MD & CEO in Leadership Overhaul

6. Apple Loses Top UI Designer as Alan Dye Joins Meta

A string of high-profile exits by senior AI leaders is emerging as one of the clearest signals of where the industry is right now. Record capital, urgent compute demand and rising internal constraints are pushing top technologists to step away from Big Tech and build on their own.

Total AI investment reached about $252 billion in 2024, with private AI investment jumping nearly 45% and generative AI alone drawing roughly $33.9 billion, according to the 2025 Stanford AI Index. 

CB Insights now estimates that global AI funding in the first half of 2025 is on track to surpass last year’s record, while US venture-market tracker PitchBook and the National Venture Capital Association say roughly half of global venture deal value this year is flowing into AI startups.

Regulators and market analysts, meanwhile, have begun warning of an AI-driven asset bubble, noting how data-center build-outs and chip purchases are increasingly being financed with debt as tech giants flood bond markets to lock in capacity.

Yet on the operational side, AI adoption within companies continues to rise, with surveys showing that a large majority of firms now use AI in their business and expect to increase budgets further.

On the numbers, this is a boom with bubble worries attached, not a slump.

That is the backdrop against which some of the most senior names in AI and software are deciding that the better way to ride this wave is to become founders again.

On 19th November, Meta’s chief AI scientist Yann LeCun confirmed in a LinkedIn post that he is leaving the company at the end of the year after 12 years, writing, “I am planning to leave Meta after 12 years: 5 years as founding director of FAIR (Facebook AI Research, Meta’s fundamental AI lab) and 7 years as Chief AI Scientist. I am creating a startup company to continue the Advanced Machine Intelligence research program (AMI) I have been pursuing over the last several years.”

After a decade helping to build Meta’s research muscle and open-source models, he is now betting that the next stage of AMI is better built outside.

A few weeks earlier, another AI veteran made a similar move. In a mid-September LinkedIn post, enterprise AI-model platform MosaicML co-founder Naveen Rao told colleagues, “Today is my last day at (data and AI cloud firm) Databricks,” before adding: “Innovation and investment in the AI era are taking on new forms. We’re spinning out a new effort around foundational challenges in AI and computing and I’m excited to be leading it.”

By early October, reports had identified the venture as Unconventional, Inc., a startup focused on the compute and systems bottlenecks created by the AI boom.

The pattern extends beyond research labs. On 11 August, GitHub CEO Thomas Dohmke posted a farewell note on the company’s blog titled “Auf Wiedersehen, GitHub,” announcing his decision to step down to become a founder again after more than a decade at Microsoft and GitHub. 

“After all this time, my startup roots have begun tugging on me, and I’ve decided to leave GitHub to become a founder again,” he wrote.

His LinkedIn now presents him as an entrepreneur building a new venture, following his exit from GitHub.

The trend first surfaced in India earlier in the AI and cloud cycle, when Sharad Sanghi, who had long led data-center provider Netmagic Solutions and NTT’s India operations, resigned in mid-2023 and founded AI acceleration cloud startup Neysa, which launched publicly in 2024.

Many investors and founders trace this surge directly to the timing and shape of the AI boom.

Abhishek Goyal, co-founder of startup and VC data platform Tracxn, argues that the current moment looks bigger than the internet and mobile cycles. 

“With AI, there are many big opportunities being unlocked. Last we witnessed the beginning of such a wide disruption was when the internet (and mobile internet) opened up many markets over 20 years ago. AI is expected to create a much larger impact and create many more white spaces. This is a once-in-a-generation era of abundant opportunity,” he says.

Ashish Bhatia, founder and CEO of startup accelerator and seed fund India Accelerator, links the founder push to what is happening inside big companies as much as outside. 

“Most large tech firms internally have reduced middle management, tightened layers, and redirected resources to core AI bets. Leadership roles are narrower, more operational, and in many ways, less entrepreneurial,” Bhatia says. 

At the same time, he notes that it is easier and cheaper to start now, with far more off-the-shelf infrastructure and hungry capital chasing AI-native products. “When industry cycles change, and the upside shifts outside the walls, leadership talent follows the opportunity.”

Ranjeet Shetye, venture partner at early-stage venture fund YourNest and MD at supply-chain risk analytics firm Everstream Analytics, sees a clear line from the ChatGPT moment in late 2022 to this year’s career moves. “The timing is definitely tied to the AI boom… This was not feasible before ChatGPT burst onto the scene.” 

In his view, AI tools reduce both the time and the manpower needed to get first products out, opening up spaces that would have been too expensive or slow even five years ago.

Funding data largely backs up their instinct. The Stanford AI Index shows total AI investment growing more than thirteenfold over the last decade, with a sharp acceleration in 2024.

CB Insights finds that AI funding has remained above $45 billion per quarter for four straight quarters through Q3 2025, even as the number of deals has fallen and money has concentrated in a handful of mega-rounds for companies such as OpenAI, Anthropic, and xAI.

PitchBook’s global venture monitor describes 2025 as a bifurcated year, with AI enjoying “exuberance and capital concentration” while many other startup sectors remain subdued.

That wall of money is now bleeding directly into executive career decisions. CEO roles themselves are turning over at record rates. CEO turnover research firm Challenger, Gray and Christmas reports that 2,221 US CEOs left their posts in 2024, up 16% from the previous record of 1,914 in 2023.

Through August, CEO exits hit 1,504, the highest on record for that point in the year and 4% above the same period in 2024.

The technology sector has seen one of the biggest increases, with CEO exits up about 6% year on year.

Most of those departures do not lead to new startups. Many CEOs retire, move onto boards or join private equity. But the churn adds to the sense that this is a once-in-a-cycle reset. In parallel, AI-heavy firms are tapping bond markets at record levels to fund data centers and chips, reinforcing the perception that we are in an “all-in” moment for AI infrastructure.

On the founder side, investors openly say they are more comfortable backing people with large-company or second-time-founder experience. 

Tracxn’s Goyal points out that “second-time founders, CXOs of large startups, executives from tech giants and founders from top colleges” draw more interest than unknown first-timers. 

India Accelerator’s Bhatia frames it as risk management, not bias. “Someone coming from a respected tech company brings a known operating environment, exposure to scaled systems, clarity on execution style, and a strong network of early hires and advisors. For a VC, this removes several layers of uncertainty.”

That does not mean a brand-name résumé is enough. YourNest’s Shetye is blunt about the filter: “Very few will possess the humility, drive, resilience, curiosity, first principles thinking, critical thinking, discipline, and coachability to successfully traverse this challenging journey and emerge victorious.” 

High AI valuations and abundant capital, in other words, do not guarantee that these new ventures will survive the next downturn.

The open question is what happens if and when the AI bubble thesis is tested. The Bank of England’s latest Financial Stability Report names AI valuations as a risk to global markets, citing the growing use of debt to fund data-center build-outs.

Analysts at Evercore, the OECD and others have started to list an AI-driven equity correction as a downside scenario for the US economy.

At the same time, Nvidia and other chipmakers insist this is a structural shift rather than a speculative spike, pointing to still-rising demand for AI compute from cloud providers.

For now, founders like LeCun, Rao, Dohmke and Sanghi are acting as if the structural story will win. Big Tech is doubling down on frontier models and trillion-dollar data center plans, while smaller teams move faster on specialized tools, vertical applications and new ways to manage compute.

As Bhatia puts it, “When industry cycles change, and the upside shifts outside the walls, leadership talent follows the opportunity.” 

His broader conclusion could serve as the thesis statement for this whole wave: when elite talent steps outside, it is usually because they believe the next big leap will come from agility, not size.

On the evidence, AI is not in a slump. It is in a capital-heavy boom with bubble warnings, forcing a generational choice on senior leaders. Stay inside the giants and help steer multi-trillion-dollar bets on models and infrastructure, or step out and try to build the next layer of the stack from scratch.