When all your cofounders leave, you're not running a startup — you're running a one-man show with 1,200 extras.
The Summary
- Every xAI cofounder except Musk has departed, along with about 80 identified employees across engineering, program staff, and legal in the past year — significant given the company only launched in 2023
- The exodus overlaps with a new org structure announced in February, a SpaceX merger, a reported $60 billion Cursor acquisition option, and an upcoming IPO
- xAI had roughly 1,200 employees as of March 2025, meaning the documented departures represent at least 6-7% of the workforce — and those are just the publicly searchable ones
The Signal
xAI is bleeding talent at exactly the moment it's trying to prove it belongs in the same conversation as Anthropic, OpenAI, and Google. The departure list reads like a DeepMind and OpenAI alumni reunion in reverse: Igor Babuschkin, Kyle Kosic, Christian Szegedy. These weren't junior engineers or mid-level PMs. These were the founding technical core, the people who were supposed to make "maximally curious" and "pro-humanity" AI something more than a press release talking point.
The timing tells you everything. February 2026: new org structure announced. Spring 2026: key people from that structure are already gone. You don't reorganize a company and immediately lose the people meant to execute the new plan unless something fundamental is broken. Either the vision changed under their feet, the execution environment became untenable, or both.
"When your cofounders leave, they're not just changing jobs. They're voting with their equity."
Now layer in the strategic chaos. xAI is simultaneously merging with SpaceX, reportedly negotiating a $60 billion option to buy Cursor (an AI coding assistant), and prepping for an IPO. That's not a roadmap. That's three different strategies fighting for oxygen in the same room. Are you building foundational models? Are you a tools company? Are you a SpaceX subsidiary optimizing rockets with LLMs? The people leaving probably couldn't answer that question either.
Meanwhile, Grok, xAI's flagship chatbot, continues to face criticism. The product isn't breaking through. It's not the default choice for developers, researchers, or even Musk's own companies beyond X. When your founder runs multiple companies and still isn't dogfooding your AI across all of them, that's a signal.
Key context:
- xAI launched in 2023 with explicit goals to compete with top-tier AI labs
- The company raised massive capital but faces a crowded market with entrenched players
- Leadership churn at this scale, this early, suggests either cultural or strategic misalignment
The agent economy runs on two things: models that work and teams that can ship them. xAI is losing the second while the jury's still out on the first. You can raise billions. You can merge with SpaceX. You can bid $60 billion for Cursor. None of that matters if the people who know how to build what you're promising keep walking out the door.
The Implication
Watch where the xAI alums land. If they cluster at competitors or start their own labs, that's the real story. Brain drain doesn't just weaken the company people leave. It strengthens the ones they join. Anthropic, OpenAI, and the next wave of AI startups just got a talent infusion of people who've seen what doesn't work at scale.
For anyone building in the agent space: this is what happens when strategy becomes noise. Pick a lane. Build something that works. Keep the people who can execute. Everything else is just headlines.