Google’s AI Leadership Shake-Up Signals a Stronger Push for Gemini

SAN FRANCISCO — Google is reorganizing its artificial intelligence leadership as it faces growing pressure from rivals and renewed urgency around its Gemini model. The changes are also reshaping the balance between Google DeepMind’s long-running research culture and Alphabet’s commercial ambitions.

At the center of the transition is Google co-founder Sergey Brin, who has become increasingly involved in AI efforts despite no longer holding an executive position. People familiar with his involvement say Brin has encouraged researchers and engineers to focus intensely on improving Gemini and closing the performance gap with competing systems.

The renewed push comes after Google briefly moved ahead of rivals with Gemini before losing ground again as competitors released newer models. Internal testing reportedly led Google to postpone the next flagship Gemini release by about two months, with coding performance among the areas where the model continued to face challenges.

Hassabis Moves Away From Daily Management

Google’s biggest leadership change came on August 5, when Demis Hassabis stepped back from his role running Google DeepMind on a day-to-day basis.

Hassabis, one of the most prominent figures in modern AI research, became chairman of DeepMind and took on a broader chief scientist role at Alphabet. His new responsibilities are expected to concentrate more heavily on long-term scientific questions and the potential impact of advanced AI.

His deputy, Koray Kavukcuoglu, has taken a more powerful operational position. Google says Kavukcuoglu will have the final say on major decisions involving DeepMind. He will also continue overseeing Gemini, a responsibility he had already assumed more than a year ago.

The transition represents more than a change in job titles. It reflects a gradual shift in how Google wants its AI organization to operate, with Gemini increasingly positioned at the heart of the company’s broader product and business strategy.

DeepMind’s Independence Faces New Questions

Another significant element of the restructuring is the movement of some teams from DeepMind into Google’s corporate organization.

The change has raised concerns among employees about whether DeepMind will retain the level of independence it historically enjoyed. During an internal meeting following the announcement, staff were told that several nontechnical groups would move into Google’s wider corporate reporting structure.

DeepMind has long maintained a strong research-oriented identity, pursuing projects that extend beyond immediate commercial applications. But Alphabet has increasingly sought to integrate AI technology into products and services that can generate revenue.

That tension has become more important as the competition for advanced AI intensifies. Google needs its research organization to produce leading models while also making those systems useful across its consumer and enterprise businesses.

Brin’s Return to the AI Push

Brin’s involvement adds another layer to the restructuring.

The co-founder has reportedly been encouraging employees to concentrate on areas that could help Google regain the technological lead. One area receiving particular attention is recursive self-improvement — AI systems capable of improving aspects of their own development with reduced human intervention.

Brin previously became more active in Google’s AI efforts after the arrival of ChatGPT shook the technology industry in 2022. His renewed involvement now comes as Google confronts another period of intense competition from companies including OpenAI and Anthropic.

Although Brin does not formally run Google’s AI organization, his position as co-founder gives his views considerable weight when strategic priorities and resources are discussed.

Kavukcuoglu Emerges as a Central Figure

Kavukcuoglu’s growing influence has been developing for some time.

Google CEO Sundar Pichai appointed him as the company’s chief AI architect in 2025 while he was still serving as DeepMind’s chief technology officer. The role was designed in part to help spread Gemini technology throughout Google’s products.

He later moved from London to Mountain View, bringing him closer to Google’s main headquarters and senior corporate leadership. Sources cited by Reuters said his influence over Gemini’s technical direction grew as Hassabis became less involved in the model’s day-to-day management.

His stronger connection with Google Cloud is also significant because the cloud business is one of Alphabet’s major commercial channels for AI technology. The restructuring could therefore help bring research, infrastructure and commercial priorities into closer alignment.

Departures Add to the Pressure

The leadership overhaul also coincided with the departure of prominent Google AI figures, including Jeff Dean and Oriol Vinyals, who had been among Gemini’s original technical leaders. They are leaving to establish a new startup.

The exits have created uncertainty around several research teams and projects that had operated with considerable independence. Employees are now watching closely to see how those groups will be reorganized and whether their work will remain focused on broader scientific research or become more closely tied to Gemini and Google’s commercial objectives.

Google’s challenge is therefore twofold: it must accelerate development of competitive AI models while maintaining the research capabilities that helped make DeepMind one of the industry’s most influential laboratories.

The latest restructuring suggests Alphabet believes a tighter organization, stronger central leadership and greater emphasis on Gemini are necessary to compete at the frontier of AI. Whether that approach can translate into faster model development and sustained technological leadership will become clearer as Google prepares its next generation of AI systems.