Google Loses Two Top AI Researchers To OpenAI & Anthropic

Search Engine Journal reports that two of Google’s highest-profile AI researchers — Noam Shazeer and John Jumper — are departing for OpenAI and Anthropic respectively, within days of each other. The article ties the exits to broader competitive pressure on Google’s AI strategy, particularly around developer tooling.

What’s actually new

Shazeer, co-author of the “Attention Is All You Need” paper that introduced the Transformer architecture, announced his move to OpenAI on June 18. Google had brought him back in 2024 through a deal with Character.AI reported at $2.7 billion, making him a co-lead on Gemini — and he’s leaving less than two years later. Jumper, who shared the 2024 Nobel Prize in Chemistry for AlphaFold, is heading to Anthropic after nearly nine years at Google DeepMind. Alphabet shares dropped roughly 5% to 6% on June 22. Bloomberg reporting cited by SEJ points to internal DeepMind concerns that Google lacks a clear product for businesses building AI coding tools — an area where Anthropic and OpenAI have been gaining traction. Google CEO Sundar Pichai himself acknowledged in May that Google was “a bit behind” on agentic coding.

What it means for your config

This is a talent and strategy story, not a product change. Nothing about the Gemini API, AI Overviews, or any Google developer-facing tool is changing as a direct result of these departures — at least not today. If you’re building on Gemini models or integrating Google’s AI APIs, your existing configurations remain valid.

That said, the subtext matters for anyone making long-term tooling bets. The Bloomberg reporting about Google lacking a clear AI coding product for businesses is worth paying attention to if you’re evaluating which AI-assisted coding platforms to build workflows around. Anthropic and OpenAI are both actively hiring from Google and shipping developer-facing products at a faster clip. If you’re deep in a Google-centric AI stack, this doesn’t mean you need to migrate tomorrow, but it’s a signal to keep your abstractions flexible and avoid tight coupling to any single provider’s SDK patterns.

No config changes, no breaking updates, no migration needed. We’ll revisit if any of these departures lead to actual product or API shifts down the line.

Recommended next step

Read the original for the full timeline and market context. If you’re evaluating AI coding tools for your team, this is a useful data point to file alongside the actual product comparisons. The concrete thing to watch is Anthropic’s AI for Science event on June 30 and whatever OpenAI ships next on the coding front — those are the moments where talent moves translate into product changes that could actually affect your stack.

---

Read the full announcement on Search Engine Journal → Google Loses Two Top AI Researchers To OpenAI & Anthropic