Noam Shazeer, Google’s vice president of engineering and one of the leaders of the company’s Gemini model project, has announced that he is leaving Google to join OpenAI. In a post on X, he said he was “excited to share” the move and called it “a difficult decision,” adding that he was proud of “the amazing team at Google” and what they built together. Google said it was grateful for his “significant contribution” over the years.
Shazeer joined Google in 2000 and later helped write a landmark 2017 research paper widely seen as a catalyst for the AI boom. He was promoted to co-lead Gemini development and was considered a central figure in improving the model’s capabilities relative to OpenAI’s ChatGPT.
This is the second time Shazeer has left Google. He first departed in 2021 with Daniel De Freitas to found the AI startup Character.AI. Both returned to Google in August 2024 after a $2.7 billion deal with the company. Character.AI let users chat with virtual versions of celebrities such as Billie Eilish or anime characters, and also create their own chatbots and AI assistants.
The move highlights the fierce battle for skilled AI researchers and engineers, which has become a key front in competition among technology companies. The announcement came weeks after Google unveiled new Gemini AI products and as OpenAI is reportedly heading toward an expected initial public offering.
