A prominent Google researcher is leaving for OpenAI, dealing another blow to Alphabet in the increasingly fierce battle for AI supremacy.

For a company that essentially invented the transformer architecture, the foundational technology behind modern AI, Google keeps watching its researchers walk out the door.

The talent tug-of-war that won’t stop

This isn’t an isolated incident. It’s part of a well-documented trend that accelerated in early 2023, when Jacob Devlin, a well-known Google AI researcher, left for OpenAI. Devlin’s departure was notable because it wasn’t just about a better offer. He reportedly had internal concerns about the data practices behind Google’s Bard chatbot.

That exit opened the floodgates. Multiple Google AI researchers followed Devlin out the door in that same period, prompting talk of a “brain drain” at what had been considered the premier destination for AI talent.

The flow hasn’t been entirely one-directional. Ben Goodger made the reverse trip in April 2026, returning to Google Labs after a stint at OpenAI.

Over 20 researchers have departed from major AI labs, including both Google DeepMind and OpenAI, for startups like Periodic Labs between 2025 and 2026.

Why each departure matters more than it seems

The transformer paper, “Attention Is All You Need,” came out of Google in 2017. Yet OpenAI has consistently managed to commercialize transformer-based technology more aggressively, turning research breakthroughs into products like ChatGPT that dominate public consciousness.

The departures also raise questions about what’s driving them. In Devlin’s case, it was ethical concerns about data practices. For others, compensation has reportedly been a factor, with OpenAI and well-funded startups offering packages that even Google’s deep pockets struggle to match.

What this means for investors and the broader market

The steady outflow of talent from established labs to both competitors and startups suggests the innovation center of gravity in AI may be shifting. Over 20 top researchers scattering to smaller firms creates a more fragmented competitive landscape.

Companies like Periodic Labs, founded by former big-lab researchers, represent potential acquisition targets or future competitors. The formation of new AI startups by ex-Google and ex-OpenAI talent has become one of the most watched trends in Silicon Valley dealmaking.

The back-and-forth nature of these moves, exemplified by Goodger’s return to Google, also suggests that neither company has a permanent advantage in retention. For a field where multi-year research projects are the norm, that instability carries real costs in terms of delayed timelines and disrupted collaboration.