Google’s biggest AI reshuffle is meant to make the company faster. Instead, it has put a spotlight on the talent, product delays and ethical friction that could make that task harder.

On Wednesday, Google moved Demis Hassabis out of day-to-day leadership of Google DeepMind, installing CTO Koray Kavukcuoglu to run the unit while Hassabis becomes DeepMind chair and Alphabet chief scientist. Hassabis said he was handing over operations to gain “the time and space to focus on the big picture.”¹ Pichai framed the change as a bid to sharpen Google’s frontier push, saying the company must “accelerate all this work and stay focused on the AI frontier.”Sundar Pichai (sundarpichai)²

The timing matters. Google has faced delays to Gemini 3.5 Pro while OpenAI and Anthropic have released newer frontier systems, particularly in lucrative coding tools. Critics outside the company read the reorganization as a loss of influence for Hassabis and a more product-driven turn; Google says the leadership change was not tied to release delays.³

Then came the more personal blow: Jeff Dean, Google’s 30th employee and a 27-year veteran, is leaving with Sanjay Ghemawat, Quoc Le and Oriol Vinyals to found Discovery Loop. The public-benefit company aims to automate entire scientific and engineering experimental cycles, running thousands of experiments in parallel. “The next great frontier for AI is to go beyond answering questions and to begin making discoveries,” its founders said.⁴ Pichai called Dean “incomparable” and said Google would support the venture.Sundar Pichai (sundarpichai)⁵

For Google employees and ethics advocates, however, the departure carries a different charge. Some link it to employee backlash over Google’s Pentagon arrangement, which permits military use of its AI, and to Dean’s public opposition to mass surveillance and autonomous lethal weapons. One current employee called his exit part of a growing brain drain: “Mediocrity is the likely result.”⁶

Google still has enormous compute, chips, cash and reach. But its reset now faces a harder test: prove that faster execution does not come at the expense of the people—and principles—that built its AI lead.