Google Loses Two AI Leaders as Transformer Co-Author Joins OpenAI, AlphaFold Father Moves to Anthropic

According to announcements on X, Google lost two influential AI researchers within days in June 2026. On June 18, Noam Shazeer, co-author of the Transformer architecture that underpins modern large language models, announced he was leaving Google to join OpenAI. Shortly after, John Jumper, who led the development of AlphaFold protein-folding system and won the 2024 Nobel Prize in Chemistry, declared he would depart DeepMind after nearly nine years to join Anthropic.

According to Bloomberg, Shazeer's exit reflects organizational challenges at large tech firms. Google DeepMind employs roughly ten thousand researchers across fragmented product lines including Gemini, Gemma, and Veo, each with separate reporting structures and priorities. In contrast, OpenAI is characterized as an "AI-native" company moving faster on hiring, investment, and product velocity. Meanwhile, Anthropic has been investing heavily in life sciences, aligning with Jumper's expertise following its $400 million acquisition of Coefficient Bio in April.