In the space of seven days last June, Google DeepMind didn’t just lose a few researchers — it lost a piece of its soul. Four of its most senior scientists walked out the door. One of them, Alex Turner, didn’t leave for a better offer or a faster GPU. He left because he couldn’t reconcile working at a company he believed was complicit in mass surveillance and immigration enforcement.
Turner’s “Why I Left” essay, published July 15, is the rare resignation letter that reads like a moral manifesto. He talks about his failed attempts to push Google to reconsider its military AI contract with the Pentagon — one with no restrictions against killer robots or mass surveillance. He tried talking to Jeff Dean, Google’s legendary AI leader. He tried rallying support from luminaries like Stuart Russell. It didn’t matter. Internal proposals were ignored. Then came the death of Alex Pretti on January 26, 2026 — an event linked to Google Cloud services used by immigration enforcement. That steeled Turner’s resolve. He left.
But Turner was only the last straw. The week before his departure, the exodus was already a crisis. On June 17, Noam Shazeer — co-author of the seminal “Attention Is All You Need” paper — left for OpenAI. He had returned to Google just two years prior via a $2.7 billion Character.AI licensing deal. Two days later, Nobel Prize winner John Jumper, the architect of AlphaFold 2, announced he was joining Anthropic. Two more senior researchers, Jonas Adler and Alexander Pritzel, followed Jumper to Anthropic. The result? Alphabet’s market cap dropped by anywhere from $50 billion to $270 billion, depending on who you ask.
Here’s the twist: this isn’t just about money or prestige. The research summary repeatedly cites bureaucratic friction, product fragmentation, and a lack of startup urgency as root causes. Google’s AI teams have become so bloated that even Nobel laureates feel they can move faster outside its walls. Anthropic and OpenAI now offer comparable compute, more autonomy, and — critically — a higher perceived chance of leading the frontier. For Jumper, Anthropic’s push into AI for science — including wet labs and partnerships with the Allen Institute — made the move a no-brainer.
Who loses here? Alphabet, obviously. But also every Google employee who believed the company could balance ethics and profit. Jeff Dean lost a battle of persuasion. The memory of Alex Pretti becomes a rallying cry for others who might now consider leaving. And the AI field loses a once-unified front against harmful applications.
Who wins? Anthropic and OpenAI, who just absorbed some of the brightest minds in AI. And the researchers themselves — Shazeer gets another payday, Jumper gets a faster lab, and Turner gets his conscience back.
What happens next? If another Nobel-level researcher leaves before August, expect another double-digit percentage drop in Alphabet shares. But the bigger question is whether Google will — or even can — change its culture fast enough. Or will the “why I left” genre become a movement that reshapes the entire AI talent landscape?