Editorial illustration for Google moves AI teams from DeepMind as focus shifts from London to California
Google Shifts AI Teams From DeepMind London to California
Google is pulling AI engineering work out of DeepMind's London operation and consolidating it in California, according to people familiar with the restructuring. The move comes as the company tries to shake off months of speculation that it had fallen behind OpenAI and Anthropic in the race to build the most capable models.
Gemini 3 launched late last year to strong reviews, but the momentum didn't last. Rivals shipped newer systems that matched or beat it within weeks, and Google's follow-up releases looked more like patchwork than progress. Jeff Dean, one of the company's most senior AI figures, left to start his own venture, taking other researchers with him. Demis Hassabis, who runs DeepMind and holds a Nobel Prize for his work on protein folding, has had to manage the fallout while defending the lab's direction.
Relocating teams from London to California signals something bigger than an org chart reshuffle. It points to where Google thinks the next round of frontier models will actually get built, and who it wants in the room when decisions get made. The timing matters too: this shift lands just as Google and Meta both resurface with new models aimed squarely at OpenAI and Anthropic.
Why this matters
The reshuffling at DeepMind tells us more about Google's priorities than any benchmark chart could. Moving teams out of London and into California isn't a logistics footnote, it's a signal that Google wants products shipping faster than Demis Hassabis's research org was built to deliver. For developers and founders building on Gemini or DeepMind's tooling, that's worth watching closely: a shift toward commercialization often means faster iteration cycles, but it can also mean research priorities get bent toward whatever sells, not whatever's most interesting scientifically.
Researchers inside Google should be asking whether this move dilutes DeepMind's original mandate, the one that produced AlphaFold and other work far removed from quarterly roadmaps. With Anthropic and OpenAI setting the pace and Meta throwing money at catching up, Google clearly decided speed matters more than its old structure. Whether that trade pays off depends on if California's "ship it" culture can coexist with the kind of patient research that made DeepMind's reputation in the first place.
Related Reading
Browse all Google news →
LLMs & Generative AI ChatGPT's 'Nerdy' tweak rewards goblin metaphors in answers, study finds
LLMs & Generative AI Google to End Free Access to Gemini AI for Non-Subscribers
LLMs & Generative AI AI Engineers Face Rising Costs, Need New Strategies for Efficiency
Related Topic NVIDIA and Google Cloud let developers scale AI from prototype to production
Market Trends
Related Topic Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks
From Archives Orion-1 links logs AI guardrails miss as they block defenders
From Archives EU sees agentic AI as next chapter to deepen use and boost productivity
From Archives OpenAI, Google AI models easily tricked into making nudes, researchers say
Research & Benchmarks
From Archives Gemini 4 Argon review: GraphWalks F1, DeepSWE, Vibe Code Bench
Research & Benchmarks
Common Questions Answered
Why is Google consolidating AI engineering work from DeepMind's London operation to California?
Google is moving AI teams from London to California as part of a strategic restructuring to accelerate product development and compete more effectively against OpenAI and Anthropic in the race to build the most capable AI models. The shift signals that Google prioritizes faster commercialization and iteration cycles over the research-focused approach that DeepMind's London operation was traditionally built to deliver.
What happened to Google's momentum after launching Gemini 3?
Although Gemini 3 launched late last year to strong reviews, the momentum did not last because rival companies shipped newer systems that matched or beat it within weeks. Google's follow-up models failed to maintain competitive advantage, contributing to months of speculation that the company had fallen behind in the AI race.
How did Google and Meta respond to competition from OpenAI and Anthropic?
Google and Meta both unveiled frontier models that they claim can compete directly with the best systems from OpenAI and Anthropic, reigniting the competitive race for the most capable AI models. This development indicates that the competition has become more balanced, with multiple companies now capable of delivering competitive frontier-level AI systems.
What does the DeepMind restructuring reveal about Google's strategic priorities?
The reshuffling of AI teams from London to California demonstrates that Google is prioritizing faster product commercialization and shipping speed over the pure research focus that characterized DeepMind's traditional organizational structure. This shift toward California-based operations suggests Google wants to accelerate iteration cycles and market responsiveness to better compete in the rapidly evolving AI landscape.
Further Reading
- EXCLUSIVE: Inside the Google executive moves that led to its big AI reshuffle - Reuters
- Google shifts AI power to California in race against Anthropic, OpenAI - The Straits Times
- Rivalry At Google DeepMind Preceded Recent Shakeup - Forbes
- Breakingviews - COMMENTARY: London AI scene can thrive despite DeepMind reset - Reuters Breakingviews
- Google moves AI responsibility team out of DeepMind - The Wall Street Journal
