Noam Shazeer Joins OpenAI from Google

According to Reuters and The Information, Noam Shazeer, a vice president of engineering at Google and co-lead of its Gemini models, announced on X that he will leave the company to join OpenAI. Reuters reports Google told the agency it is "grateful for Noam's meaningful contributions" and that the timing of the departure was not immediately clear. Coverage across Reuters, CNBC, and The Information notes Shazeer joined Google in 2000, left to cofound Character.AI, then returned to Google in 2024 under a deal public reporting describes as involving $2.7 billion, a figure reported by The Wall Street Journal and cited by Reuters. Business Insider published excerpts of his X post, including: "I'm excited to share that I'll be joining OpenAI and look forward to working with the exceptional team there."

Reuters and The Information report that Shazeer co-authored the 2017 paper "Attention Is All You Need," which introduced the transformer architecture that underpins virtually every major large language model developed since. For practitioners, his departure to OpenAI carries more weight than a typical engineering-executive move because of that foundational research background. Engineers with deep architecture knowledge, training-pipeline experience, and large-scale product launch history at a rival lab can accelerate model capability improvements and evaluation rigor at the receiving organization.

Reuters, CNBC, and The Information frame the move as part of an ongoing and escalating talent competition among major AI labs. Public reporting of large financial transactions tied to personnel, such as the $2.7 billion acqui-hire cited by Reuters and the Wall Street Journal, underscores how firms use acquisitions and compensation packages to secure teams and implementation knowledge. Shazeer's announcement follows a period in which OpenAI, Google DeepMind, and other top labs have each made high-profile hires and departures.

Observers will track OpenAI's research publications, preprint filings, and benchmark submissions for evidence of new architectural directions or optimization patterns. Analysts will also watch Google DeepMind's Gemini roadmap and engineering team announcements for signs of gap-filling hires. Reporting so far does not specify Shazeer's role at OpenAI beyond his X post, and Google has issued only a brief statement of gratitude, per Reuters.

High-profile talent movements at leading labs have historically preceded shifts in training infrastructure, evaluation methods, and scaling strategies. Teams should monitor upstream repositories, preprint servers, and benchmark leaderboards to track where engineering knowledge is consolidating.