Noam Shazeer, one of the most influential engineers in modern artificial intelligence, is leaving Google to join OpenAI, a move he announced on June 18, 2026. Shazeer served as a vice president of engineering at Google and a co-leader of its Gemini models, placing him at the center of the company’s effort to compete at the frontier. His departure for OpenAI ranks among the most significant individual moves in an industry where a single researcher can shape the direction of an entire lab.
Shazeer’s name carries unusual weight because of what he helped create. He is a co-author of the 2017 paper “Attention Is All You Need,” which introduced the Transformer, the architecture that underpins nearly every large language model in use today, including the systems built by OpenAI, Google, and Anthropic. Engineers who laid the groundwork for the current wave of artificial intelligence are rare, and the companies that employ them treat them as strategic assets rather than ordinary hires.
The move is a sharp reversal for Google. The company paid roughly $2.7 billion in 2024 to bring Shazeer and part of his team back from Character.AI, the startup he had left to build, in a deal that was widely read as Google buying back talent it could not afford to lose. Two years later, that investment walks out the door to the chief rival in the model race, a sequence that highlights how hard it is to hold onto the small group of people capable of pushing the technology forward.
Shazeer framed the decision as difficult but forward looking. “I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there,” he wrote in a post on X, adding that he was proud of what his team had built at Google. The measured tone did little to soften the signal it sent, since the engineer who co-led Gemini is now moving to the company whose models Gemini was built to beat.
OpenAI chief executive Sam Altman welcomed the hire in unusually personal terms, calling Shazeer “one of the people I have most wanted to work with since the very beginning of OpenAI.” For Altman, landing a researcher of Shazeer’s standing is both a practical gain in engineering depth and a statement about where top talent believes the most interesting work is happening. The competition now playing out across the AI model race is happening in hiring as much as in benchmarks.
The timing sharpens the stakes. Shazeer departs Google as the assistant market fragments and as ChatGPT’s long held lead has slipped, a shift that has put every major lab under pressure to ship stronger models faster. Losing a co-lead of Gemini at that moment forces Google to reassure both its own engineers and the outside world that its roadmap does not depend on any single person. For OpenAI, the hire adds firepower at exactly the point when it is racing to defend its position against Gemini, Claude, and a widening field.
Talent has become one of the clearest battlegrounds in artificial intelligence, and moves like this one ripple well beyond a single org chart. A researcher who helped invent the foundation of modern models brings not just skill but credibility, and the labs know that where the best people go, others tend to follow. Whether Shazeer’s arrival reshapes OpenAI’s research or simply strengthens an already deep bench will take time to show, but the message to the rest of the industry landed the moment he posted it. The race for the strongest model is, at its core, a race for the people who can build it.
Related: Anthropic’s Claude Tag teammate in Slack · OpenAI and Anthropic’s next models
The big picture: The AI Model Race (2026): who’s winning the frontier contest
Your First 10 AI Skills
10 practical AI skills, copy-paste prompts and a 7-day plan to start using AI with confidence.
Download the guide →