Google Hires Former OpenAI Executive Barret Zoph

Google has hired Barret Zoph as its new vice president of research, marking a return to the search giant for the prominent artificial intelligence researcher. A Google spokesperson confirmed the appointment, noting that Zoph will apply his expertise in reinforcement learning and post-training to the development of Google's Gemini model family.

Zoph's career path highlights the rapid executive turnover currently defining the competitive AI landscape. After an initial stint at Google, Zoph spent two years at OpenAI. He departed the high-profile AI lab in October 2024 to co-found the startup Thinking Machines alongside former OpenAI executive Mira Murati. However, his tenure there was brief; in January, Zoph and fellow co-founder Luke Metz departed Thinking Machines to return to OpenAI, with later reports revealing that Zoph had actually been fired from the startup.

Upon his return to OpenAI, Zoph was tasked with leading the company's AI enterprise sales division. This second stint lasted only five months before he departed the startup once again in June. His move back to Google represents his third major career transition in less than a year, reflecting a broader trend of high-level talent shifting between major tech firms and startups as the industry prepares for massive commercial scaling.

For AI practitioners and developers, Zoph's return to Google signals a concerted effort by the company to bolster its post-training pipeline for Gemini. Reinforcement learning and post-training alignment are critical phases in making large language models safer, more instruction-compliant, and commercially viable. By securing top-tier talent with deep experience in these specific domains, Google aims to close the capability gap with rivals like OpenAI, potentially leading to more robust and reliable Gemini APIs and enterprise tools in the near future.