Google and Alphabet are shifting leadership at the top of their flagship AI unit, Google DeepMind. In a major announcement, Alphabet CEO Sundar Pichai revealed structural changes designed to accelerate the company’s artificial general intelligence (AGI) roadmap while keeping its consumer AI efforts moving at top speed.
Under the new plan, Demis Hassabis—co-founder and longtime CEO of Google DeepMind—is stepping into a broader, high-level role as Chair of Google DeepMind and Chief Scientist of Alphabet, while continuing to lead Isomorphic Labs. Taking over day-to-day operations and research execution is Koray Kavukcuoglu, who has been promoted to Senior Vice President (SVP) of Google DeepMind.
Here is what these leadership moves mean for Google, Gemini’s development, and the ongoing race toward AGI.
Demis Hassabis Shifts Focus to AGI & Frontier Science
Demis Hassabis is transitioning from direct, daily management to overseeing high-level AI strategy, frontier research, and scientific discovery across Alphabet.
As Chair of Google DeepMind and Alphabet’s Chief Scientist, Hassabis will focus on:
Shaping the Future of AGI: Defining long-term strategy for artificial general intelligence beyond today’s Large Language Models (LLMs).
Scientific Discovery: Applying AI to complex real-world challenges in biology, chemistry, and materials science.
Isomorphic Labs: Continuing to lead Alphabet’s drug-discovery venture, which builds on AlphaFold technology to transform medicine.
Sundar Pichai noted that freeing Hassabis from day-to-day operational tasks allows him to focus on work that is vital for both Alphabet and the broader scientific community.
Koray Kavukcuoglu Takes Over Model Development and Gemini
Taking full charge of model development and research execution is Koray Kavukcuoglu, a 13-year DeepMind veteran who previously served as VP of Research.
As SVP of Google DeepMind, Kavukcuoglu now directly manages:
Model Research & Development: Guiding the engineering teams behind next-generation foundational models.
Gemini App & Developer Ecosystem: Connecting research breakthroughs straight into consumer apps, API features, and enterprise tools.
Who is Koray Kavukcuoglu?
Kavukcuoglu is a pioneer in modern deep learning. Over his 13 years at DeepMind, he helped build the lab’s core technical foundation and led several milestone projects, including:
Deep Q-Networks (DQN): The breakthrough that combined deep learning with reinforcement learning, enabling AI to master complex tasks like gaming directly from raw visual data.
WaveNet: The generative voice model that set a new benchmark for natural-sounding text-to-speech across the tech industry.
Deep Learning Teams: Building DeepMind’s original deep learning division from the ground up.
Why Google is Making This Move Now
This re-organization directly addresses two complementary needs: rapid commercial deployment and long-term scientific breakthroughs.
Faster Commercial Pipeline: By grouping model development, research, and the product teams behind Gemini models and apps under Kavukcuoglu, Google can ship breakthroughs to users faster with fewer organizational layers.
Dedicated Long-Term Vision: Assigning Hassabis to focus on grand-scale AGI and scientific problems keeps Alphabet positioned for the next major paradigm shift in artificial intelligence—a vision Google recently showcased during its major announcements at Google I/O 2026.
What This Means for Gemini’s Future
For users and developers, this shift signals an aggressive product push.
With Kavukcuoglu leading model development and the Gemini team under one umbrella, expect faster updates across consumer platforms, on-device mobile AI integrations, and Google Cloud services. Meanwhile, Hassabis’s focus as Chief Scientist suggests Google is actively laying the groundwork for what comes after current transformer architectures.
To stay fully up to date on all model releases, leaks, and technical breakdowns, follow our comprehensive Google AI News hub on NPowerUser.
