In short
Google is reshaping its AI leadership, moving Demis Hassabis into a chairman and chief scientist role while promoting Koray Kavukcuoglu to run DeepMind’s operations. Jeff Dean is also leaving Google to launch an AI-focused public benefit corporation.
- Demis Hassabis is stepping back from direct DeepMind leadership to become chair and Alphabet chief scientist.
- Koray Kavukcuoglu is taking over operational leadership of Google DeepMind while remaining Google’s chief AI architect.
- Jeff Dean and Sanjay Ghemawat are leaving Google to start Discovery Loop, an AI-focused public benefit corporation.
- Google will invest as a founding backer in Discovery Loop, keeping ties to the new venture.
- The changes suggest Google is reorganizing to balance frontier research, product execution and long-term AI strategy.
Google has reorganized the top of its artificial intelligence leadership, moving Demis Hassabis out of day-to-day control of Google DeepMind so he can serve as the lab’s chair and Alphabet’s chief scientist. The change matters because it shifts one of the company’s most prominent AI figures into a broader strategic role while elevating longtime DeepMind executive Koray Kavukcuoglu to lead the organization’s operational side.
The leadership reset, announced Wednesday by Sundar Pichai, also includes a separate departure: Jeff Dean, one of Google’s most celebrated technical leaders, is leaving Google with fellow engineer Sanjay Ghemawat to launch a public benefit corporation focused on AI for science and engineering.
Taken together, the moves suggest Google is tightening its AI governance structure at a moment when the company is racing to commercialize products, sharpen its research edge and defend its position against rivals across the industry.
What changed in Google’s AI leadership?
Google is splitting some of the responsibilities that had been concentrated around Hassabis, the high-profile co-founder of DeepMind and one of the most influential figures in modern AI. Under the new arrangement, he will no longer run the lab in the same hands-on capacity he has held in recent years.
Instead, Hassabis will become chair of Google DeepMind and Alphabet’s chief scientist. In practical terms, that places him in a more advisory and strategic position, with a mandate to focus on the company’s most ambitious technical bets rather than the day-to-day management of the organization.
At the same time, Koray Kavukcuoglu, who previously served as DeepMind’s chief technology officer, is moving into the role of senior vice president of DeepMind at Google. He will report directly to Pichai and remain Google’s chief AI architect, giving him a powerful central role in how the company builds and deploys AI systems.
How the new structure divides responsibility
The reorganization appears designed to separate long-range scientific direction from operational execution. Hassabis will remain involved in major AI strategy and AGI discussions, but Kavukcuoglu will be the executive accountable for the lab’s daily leadership and its interface with Google’s broader product and engineering organization.
This is a common pattern in large technology companies as AI becomes more commercially important: the public-facing visionary stays involved in roadmap-setting, while a trusted operator handles execution, staffing and cross-company coordination.
Why does Hassabis’s new role matter?
Hassabis has been one of the most visible advocates for the idea that artificial intelligence should be used to advance science and health. His move to chairman and chief scientist underscores that Google wants him focused on the long game: foundational research, scientific discovery and the company’s highest-stakes bets on advanced AI.
That matters because Google DeepMind is not just another research group. It is one of the company’s central engines for models, product features and scientific applications. Any leadership change there can affect how quickly Google translates research breakthroughs into consumer and enterprise products.
“I’ve always believed the No. 1 application of AI should be to improve human health,” Hassabis told staff, adding that AI should prove its value by helping cure diseases such as cancer.
He also said he would keep working closely with Pichai on strategic and global AGI issues while advising Kavukcuoglu and other DeepMind leaders. That suggests Google is not sidelining Hassabis; it is repositioning him where the company believes he can have the greatest leverage.
What this says about Google’s AI priorities
The restructuring signals that Google wants to keep one of its best-known AI figures close to the center of power without making him responsible for every operational decision. That can be especially important in an era when product launches, model safety, infrastructure planning and regulatory scrutiny all demand sustained attention.
It also reinforces the company’s message that AI is not merely a product feature but a strategic platform spanning consumer tools, scientific research and long-term platform development.
Who is Koray Kavukcuoglu?
Koray Kavukcuoglu is a veteran DeepMind leader who has helped shape the lab’s technical direction for years. As former chief technology officer, he has been deeply involved in the systems, models and engineering practices that underpin Google DeepMind’s work.
With this promotion, he becomes the executive most directly responsible for running DeepMind inside Google, while also continuing in his role as chief AI architect for the company. That combination gives him unusual reach across research and product implementation.
Because he now reports to Pichai, Kavukcuoglu’s elevation can also be read as a sign that Google wants tighter integration between its AI organization and the company’s top corporate leadership.
Why his appointment is strategically important
Google has spent years trying to merge the culture of a research lab with the pace of a product company. Putting a technically accomplished insider in charge of DeepMind’s operations may help bridge that gap.
It also reduces the risk of fragmentation in a business that now spans large language models, consumer assistants, enterprise tools and science-focused projects. A single leader with a deep technical background can help maintain coherence while translating research into deployable systems.
Jeff Dean is leaving to start a new AI company
The other major change is the departure of Jeff Dean, one of Google’s most storied engineers and the company’s 30th employee. He is leaving alongside Sanjay Ghemawat to create Discovery Loop, an AI-focused public benefit corporation.
Google will be a founding investor in the new venture, according to Pichai, which suggests the company wants to stay connected to the effort even as the founders move outside the Google structure.
Dean’s exit is notable because he has long been associated with Google’s deepest technical work across infrastructure, large-scale systems and machine learning. His move marks the end of another era for a company that has relied heavily on elite engineering talent to shape its AI strategy.
Discovery Loop says it aims to build AI systems that automatically address important problems in machine learning, science and engineering, and to speed up the pace of discovery so scientific and technical benefits reach the world faster.
What is Discovery Loop trying to do?
Discovery Loop is presenting itself as a company built around AI for scientific and engineering breakthroughs rather than consumer chatbots alone. Its stated mission is to develop systems that can solve hard problems and accelerate discovery across technical fields.
That framing places it in a growing category of AI ventures aiming to use automation and advanced models to help researchers, engineers and scientists work faster. By setting up as a public benefit corporation, the company is signaling that it wants to balance commercial aims with a broader social mission.
Why is Google making these changes now?
Google’s leadership shakeup comes at a moment when the AI market is moving quickly and the stakes inside the company are rising. Google DeepMind sits at the intersection of research, infrastructure and product development, which makes leadership clarity especially important.
As AI becomes central to search, productivity tools, developer platforms and scientific research, Google needs a structure that can support both rapid execution and long-term innovation. The company is also under pressure to keep pace with competitors while making its systems safer, more capable and more commercially useful.
Shifting Hassabis into a chief scientist role can help preserve his influence on the company’s scientific direction without overloading him with operational responsibilities. At the same time, promoting Kavukcuoglu gives Google a technical operator who can focus on execution across the DeepMind organization.
What the move means for Alphabet
Because Hassabis is now Alphabet’s chief scientist as well as DeepMind chair, the change extends beyond one research lab. Alphabet appears to be formalizing a leadership layer dedicated to top-level scientific strategy, not just product roadmaps.
That could matter for areas such as drug discovery, advanced model development, agentic systems and other moonshot-style projects that need patience, capital and institutional support.
It also reinforces the idea that Alphabet views AI not as a single business line but as a corporate-wide capability that must be coordinated across multiple units.
How does this fit Google’s broader AI strategy?
Google has been reorganizing around AI for years, but the latest changes make that effort more explicit. The company is trying to balance three pressures at once: pushing frontier research, shipping products and proving that AI can create real-world value.
DeepMind remains one of Google’s most important assets in that effort. It contributes model innovation, scientific applications and technical credibility. Keeping Hassabis close while giving Kavukcuoglu operational control suggests Google is trying to preserve both vision and speed.
The company’s decision to keep Dean’s new venture within its investment orbit also shows how Google may be adapting to a competitive talent environment: by allowing senior leaders to spin out while maintaining strategic ties where possible.
What to watch next
Over the coming months, the key questions will be whether the new structure produces faster product delivery, tighter coordination between research and engineering, and clearer leadership around Google’s most ambitious AI projects.
Analysts and industry observers will also watch how much freedom Hassabis has to pursue scientific initiatives, how Kavukcuoglu defines his leadership style, and whether Discovery Loop becomes a meaningful player in AI-for-science research.
For now, the message from Google is clear: its AI leadership is being refashioned for a new phase, one in which technical excellence, strategic focus and organizational discipline all matter more than ever.
Key people and roles at a glance
| Person | New role | Previous role | Why it matters |
|---|---|---|---|
| Demis Hassabis | Chair of Google DeepMind; chief scientist at Alphabet | Leader of Google DeepMind | Moves into a broader strategic and scientific role |
| Koray Kavukcuoglu | SVP of DeepMind at Google; chief AI architect | DeepMind CTO | Takes operational leadership of DeepMind |
| Jeff Dean | Co-founder of Discovery Loop | Chief scientist of Google DeepMind | Leaves Google to launch an AI-for-science venture |
| Sanjay Ghemawat | Co-founder of Discovery Loop | Google Fellow | Joins Dean in building the new public benefit corporation |
Timeline of the shakeup
- Wednesday: Sundar Pichai announces the leadership changes.
- Immediately: Hassabis transitions to chair of Google DeepMind and Alphabet chief scientist.
- Immediately: Kavukcuoglu becomes senior vice president of DeepMind at Google and continues as chief AI architect.
- In parallel: Jeff Dean and Sanjay Ghemawat leave Google to launch Discovery Loop.
- Next phase: Google and Alphabet adapt their AI strategy around the new leadership map.
The reshuffle reflects a company trying to preserve its research heritage while reorganizing for the next stage of AI competition. For Google, that means elevating technical leaders, sharpening accountability and keeping its most important AI minds aligned with its long-term goals.
Frequently asked questions
What changed in Google’s AI leadership?
Google changed its AI leadership by moving Demis Hassabis into a chair and chief scientist role, promoting Koray Kavukcuoglu to run DeepMind’s operations, and allowing Jeff Dean to leave and start a new AI company. The shift is meant to sharpen strategy and execution.
Why is Demis Hassabis stepping down from day-to-day leadership?
Demis Hassabis is stepping down from direct management so he can focus on broader scientific strategy, Alphabet-level AI direction and long-term AGI issues. He remains deeply involved, but in a role that emphasizes vision and research rather than operational control.
Who is Koray Kavukcuoglu?
Koray Kavukcuoglu is a longtime DeepMind executive and former chief technology officer who is now becoming senior vice president of DeepMind at Google. He will oversee the lab’s operations and continue serving as Google’s chief AI architect.
What is Discovery Loop?
Discovery Loop is a new AI-focused public benefit corporation founded by Jeff Dean and Sanjay Ghemawat. The company says it will build AI systems that help solve problems in machine learning, science and engineering and speed up scientific discovery.
Will Google invest in Discovery Loop?
Yes. Google will be a founding investor in Discovery Loop, according to Sundar Pichai. That keeps Google connected to the new company even as two veteran leaders leave to build it independently.
