NextFin News - The artificial intelligence arms race has moved from the cloud to the cobblestones of London, as Anthropic and OpenAI lead a massive leasing wave that has effectively exhausted the supply of prime office space in the capital. Anthropic confirmed this week it has secured 158,000 square feet at One Triton Square near Euston, a deal with British Land and Royal London Asset Management that provides capacity for up to 800 employees. The move comes just days after OpenAI signed for 88,500 square feet in King’s Cross, marking its first permanent U.K. headquarters.
The scale of the expansion is staggering for a sector that, until recently, operated out of modest satellite offices. Anthropic’s new footprint allows it to quadruple its current U.K. workforce of 200. This surge is part of a broader "tech rush" that has seen AI firms lease nearly 400,000 square feet of London office space in less than a month. Databricks also recently quadrupled its presence with a 137,000-square-foot headquarters in Fitzrovia, underscoring a collective bet on London as the primary research and commercial hub for AI outside the United States.
Pip White, head of EMEA north at Anthropic, noted that the U.K.’s unique blend of technical talent and its specific focus on AI safety issues made the expansion a strategic necessity. However, this influx of capital-rich tenants is colliding with a market under severe supply pressure. Data from Knight Frank and Colliers indicates that vacancy rates for "Grade A" or prime office space in the City and West End cores have fallen below 1%, a historic low compared to the long-term average of 7% to 8%. The scarcity is driving prime rents in the West End toward £182.50 per square foot, a level that threatens to price out non-tech sectors.
While the leasing spree suggests a vote of confidence in London’s post-Brexit economy, some analysts urge caution. Mark Stansfield, a senior director at CoStar Group who has tracked the London market for over a decade, suggests that while the AI sector is currently the "only game in town" for large-scale requirements, the concentration of demand in a single, volatile industry carries inherent risks. Stansfield, known for his data-driven and often conservative outlook on commercial real estate cycles, argues that the current frenzy mirrors previous tech bubbles where rapid physical expansion preceded market corrections. His view is not yet the consensus among more bullish real estate agents, but it highlights the vulnerability of a market where supply is so tightly constrained.
The sustainability of this growth depends on more than just venture capital. OpenAI’s recent decision to pause a major U.K. data center project serves as a reminder that physical office space is only one component of the AI infrastructure. If the underlying compute power or regulatory environment shifts, these massive leases could become liabilities. For now, the competition for talent remains the primary driver. By anchoring themselves in the "Knowledge Quarter" near Euston and King’s Cross, these firms are positioning themselves within walking distance of Google’s DeepMind and the Alan Turing Institute, effectively creating a geographic moat in the battle for the world’s most specialized engineers.
Explore more exclusive insights at nextfin.ai.
