What is the 'London Rule' in the context of AI talent migration?

The 'London Rule' refers to the observation that top AI researchers at Google DeepMind's London office are leaving because the center of gravity for pre-training and critical AI work is shifting to California, making it necessary to relocate to follow the most important scientific opportunities.

Why are researchers leaving Google for Anthropic specifically?

Researchers are attracted to Anthropic due to the potential for massive pre-IPO equity gains, as the company is valued near $965 billion and may list publicly soon, alongside a culture that is more focused and less bureaucratic than Google's.

How much did Google's stock drop when researchers started leaving?

Alphabet's shares fell as much as 7.2% intraday following the announcement of John Jumper's departure, marking the steepest drop since February and signaling investor concern over the talent exodus.

Is money the only reason AI researchers are leaving Google?

No, while financial incentives like pre-IPO equity are a major factor, researchers like Noam Shazeer and John Jumper have also cited frustration with Google's red tape, risk-averse culture, and the perception that the most significant scientific work is moving elsewhere.

The London Rule: Why Top AI Researchers Are Leaving Google for Anthropic and OpenAI

Top AI researchers are leaving Google for rivals like Anthropic and OpenAI, driven by a structural shift in pre-training work to California and the lure of massive pre-IPO equity. This trend, dubbed the 'London Rule,' explains why DeepMind's London team is seeing a mass exodus.

The Geographic Shift: The "London Rule" Explained

The core of this talent drain is structural rather than purely cultural. Analysts and insiders, including researcher Lucas Beyer, have identified a clear pattern: the "center of gravity" for pre-training—the foundational phase where models ingest massive datasets—is migrating from the UK to the US. This shift creates a specific dilemma for long-time Londoners.

When the most scientifically significant work moves to Silicon Valley, researchers face a choice: stay in London and work on peripheral projects, or relocate to follow the compute. The numbers suggest the latter is becoming the dominant trend. A 2025 analysis from venture firm SignalFire revealed that DeepMind engineers were roughly eleven times more likely to leave for Anthropic than the reverse, turning individual exits into a powerful current of talent migration.

The Financial Logic: Pre-IPO Equity vs. Mature RSUs

While cultural friction plays a role, the financial calculus of the AI boom is undeniable. At Google, a mature public company with a market cap exceeding $4 trillion, compensation is primarily tied to RSUs (Restricted Stock Units). These are valuable but offer predictable, relatively stable growth. In contrast, joining a private giant like Anthropic or OpenAI offers a different equation: pre-IPO equity.

For elite talent, the potential for outsized wealth creation is the primary draw. With Anthropic recently raising capital at a valuation near $965 billion and weighing a public listing as soon as this fall, early employees stand to gain significantly if the company's stock appreciates upon going public. The allure is simple: moving from a mature tech giant to a pre-IPO leader is the most reliable path to becoming a tech mogul in the current cycle.

Cultural Friction: Red Tape vs. Scientific Opportunity

Beyond geography and money, the nature of the work itself is driving researchers away. High-profile departures like Noam Shazeer have cited frustration with Google's "red tape" and risk-averse culture. Shazeer, who already made hundreds of millions from his previous startup exit, returned to Google only to leave again, suggesting that money alone wasn't the issue for his initial departure.

According to reporting, Shazeer feared the internet giant lacked the agility to compete with nimble startups in the AI race. This sentiment is echoed by Nobel Prize winner John Jumper, who reportedly wouldn't have left for Anthropic unless he believed the scientific opportunity was genuinely superior. The departure of Jumper and Shazeer highlights a critical problem for Google: its structure may be too slow for the pace of modern AI innovation.

Key cultural and operational factors driving these exits include:

Bureaucratic Bottlenecks: Frustration with internal red tape that slows down research and deployment.
Risk Aversion: A corporate culture that hesitates to take the bold risks required to lead the AI frontier.
Resource Allocation: Computing power being shifted to larger, centralized teams in London or Mountain View, leaving local researchers with fewer resources.
Focus Drift: Google's many priorities versus the razor-focused mission of competitors like Anthropic and OpenAI.

The Market Reaction: A Talent War Google Is Losing Ground In

The financial markets have interpreted these departures as a verdict on Google's competitive position. The 7.2% intraday drop in Alphabet stock following Jumper's exit was the steepest decline since February, reflecting investor anxiety that Google is losing the talent war. When a stock moves so drastically on the departure of a single researcher, it indicates the market priced the company for a level of dominance it now fears is eroding.

While DeepMind CEO Demis Hassabis has pushed back, calling the market "ferociously competitive" and insisting Google still wins its share of top people, the data suggests otherwise. With four senior departures in two weeks and a clear geographic mismatch between where the work is (California) and where the talent is (London), the competition has become undeniably geographic. Right now, the map is pointing away from London and toward the rivals who are winning the race for both compute and equity.