Google Loses Gemini Co-Lead Noam Shazeer to OpenAI

Noam Shazeer, one of the most influential researchers in artificial intelligence, has left Google to join OpenAI. The move marks a dramatic talent shift in the race to build the most capable AI systems on the planet and sends a clear signal about how intense the competition for top researchers has become.

A Transformer Pioneer Switches Sides

Shazeer co-authored the landmark 2017 paper "Attention Is All You Need," which introduced the transformer architecture. Transformers are a type of neural network design that allows AI models to process and generate text with remarkable accuracy. They are the foundation of virtually every major large language model, or LLM, used today. These are the AI systems that power tools like ChatGPT, Gemini, and Claude. At Google DeepMind, Shazeer served as co-lead on Gemini, the company's flagship AI model. He will now lead architecture research at OpenAI, according to sources familiar with the matter.

His departure comes less than two years after Google spent roughly 2.7 billion dollars to bring him back through the Character.AI acquisition in August 2024. That deal returned both Shazeer and fellow researcher Daniel De Freitas to Google's DeepMind AI unit. Google clearly viewed the investment as critical to staying competitive in the fast-moving AI landscape, making this departure all the more striking.

What This Means for the AI Industry

The hire is a significant win for OpenAI and a clear blow to Google. Shazeer brings deep expertise in model architecture, the core engineering discipline that determines how well an AI system learns and reasons. Model architecture is what separates a mediocre chatbot from a breakthrough research tool. For OpenAI, adding someone who literally helped invent the technology behind modern AI could accelerate progress on next-generation models.

The move also highlights how fierce the AI talent war has become. Top researchers can command enormous compensation packages, sometimes worth hundreds of millions of dollars, and often choose employers based on research freedom and the chance to shape the future of the field. Shazeer posted on X that it was a difficult decision to leave and praised his Google colleagues for the work they built together.

For Google, losing a researcher of this caliber raises questions about retention and morale inside DeepMind. The company has already seen several high-profile departures in recent years as competitors aggressively recruit top AI minds. Whether Google can replace Shazeer's contributions to Gemini will be one of the key questions to watch in the months ahead.