Did Noam Shazeer really leave Google DeepMind for OpenAI?
2026-06-19
5 AI models
1 contradiction
13 sources
ConsensusGemini 3.1Pro
Models consulted
OpenAI
Mistral
Anthropic Claude
Google Gemini
Grok
Consensus answer
Yes, Noam Shazeer left Google and its DeepMind unit to join OpenAI. He announced his departure on June 18, 2026.cnbc
At Google, Shazeer served as a Vice President of Engineering and was a co-lead for the company's Gemini AI models.investingcnbc At OpenAI, he took on the role of lead for architecture research.x
Key Context and Background:
AI Pioneer: Shazeer originally joined Google in 2000. He is widely recognized in the artificial intelligence field as one of the co-authors of the landmark 2017 research paper "Attention Is All You Need," which introduced the Transformer architecture and laid the foundation for modern generative AI.xcalcalistech
Second Departure: This move marked Shazeer's second time leaving Google. He previously left in 2021 alongside Daniel De Freitas to co-found the AI startup Character.AI. The pair returned to Google in August 2024 after Google absorbed the startup's team in a reported $2.7 billion deal.theinformationcalcalistech
Reactions to the Move:
OpenAI: CEO Sam Altman welcomed Shazeer, stating he was one of the people he had most wanted to work with since OpenAI's early days.neowin
Google: The company officially acknowledged his departure, releasing a statement expressing gratitude for his "meaningful contributions to Google over the years".investingcalcalistech
Differences
1 notable difference (1 contradiction)
across 5 models.
contradiction
The specific date of the announcement.
The announcement occurred on June 18, 2026.
Anthropic, Grok, Mistral
“June 18, 2026”
The announcement occurred on June 17, 2026.
OpenAI
“Wednesday, June 17, 2026”
How to verify: Check the calendar for June 2026 to see which date fell on a Wednesday.
5 AI models
answered this question independently on 2026-06-19. A judge from a different model family
then cross-checked the answers, scored how far they agree and flagged where they differ.
AI models can make mistakes – verify important information against the sources above.
