Google’s AI Shake-Up Raises Questions About DeepMind, Gemini, and the Race Toward AGI
Google’s artificial intelligence ambitions appear to be entering a critical new phase, as major leadership changes and fresh reports about its Gemini models raise questions about the company’s position in the global AI race.
For years, Google has been seen as one of the original powerhouses behind modern AI. Its research helped lay the foundation for large language models, and its DeepMind division has long been associated with breakthrough work in machine learning. But despite that history, recent developments suggest the company is under growing pressure from faster-moving rivals.
Reports have claimed that Google’s next-generation frontier AI models have faced repeated delays, with Gemini 3.5 Pro allegedly falling short of internal performance expectations. More recently, Gemini 3.6 Flash was released to a mixed reception, adding to concerns that Google’s AI roadmap may not be moving as smoothly as expected.
The biggest shift, however, comes from inside the company’s leadership structure.
Demis Hassabis, the CEO of DeepMind, is stepping down from the top executive role at the AI division. He will now serve as Chair of DeepMind and Chief Scientist of Alphabet, Google’s parent company. While this keeps Hassabis in a highly influential scientific position, the move marks a significant change at one of the most important AI organizations in the world.
At the same time, Jeff Dean, one of Google’s most important technical leaders and a key figure in shaping the company’s AI strategy over the past decade and a half, is leaving the company. Dean, along with longtime Google researcher Sanjay Ghemawat, is launching a new public benefit corporation called Discovery Loop. The new venture will reportedly focus on advancing artificial general intelligence, commonly known as AGI.
Alphabet CEO Sundar Pichai has said Google will be a founding member and cloud partner of Discovery Loop, suggesting that the relationship between the new company and Google may remain close. Even so, the departure of such a central figure is likely to fuel debate about Google’s ability to retain top AI talent at a time when competition for researchers and engineers is more intense than ever.
The timing is especially sensitive. Google is facing increasing scrutiny over whether Gemini can compete with the strongest AI models on the market, particularly in areas such as coding, reasoning, multimodal performance, and enterprise productivity.
According to reports, Google is also in talks with San Francisco-based AI startup Mechanize to improve the coding capabilities of Gemini. The potential deal could be worth up to $1.5 billion and may include hiring some of Mechanize’s key team members. If completed, it would show how aggressively Google is trying to strengthen Gemini’s performance in software development, one of the most important battlegrounds in AI.
The situation becomes even more interesting when considering Google’s broader AI partnerships. Apple has reportedly been paying Google around $1 billion per year to use or adapt Gemini technology for its own Apple Foundation Models. Now, with Google potentially spending even more to strengthen Gemini’s coding abilities through an outside startup, the AI ecosystem is beginning to look increasingly circular and complicated.
For Google, the stakes are enormous. The company played a foundational role in the development of transformer-based AI models, the architecture that helped power the current generative AI boom. Yet the perception has grown that Google moved too cautiously while competitors gained public momentum.
That cautious approach may have protected its core search and advertising businesses for a time, but it also allowed other AI companies to capture the spotlight. Now Google is trying to prove that it can still lead, not just contribute from behind the scenes.
The leadership changes at DeepMind, Jeff Dean’s departure, the creation of Discovery Loop, and the reported effort to boost Gemini through external AI talent all point to one conclusion: Google is reorganizing for a much more intense phase of AI competition.
Whether this marks the beginning of a stronger comeback or a sign of deeper internal challenges remains to be seen. But one thing is clear: the race for artificial general intelligence is accelerating, and Google can no longer rely solely on its legacy as an AI pioneer. The company now has to show that Gemini, DeepMind, and Alphabet’s broader AI strategy can still compete at the very top.
