AI Is Quietly Becoming London’s Biggest Office Tenant.
Quick Reads
- AI companies have leased over 375,000 square feet of London office space in recent weeks, with Anthropic and OpenAI among the biggest movers
- Anthropic signed a 158,000 sq ft lease at British Land’s One Triton Square development near Euston.
- British Land raised its earnings guidance for both 2026 and 2027 on the back of accelerating AI tenant demand, reporting 12 per cent campus rental growth and 6 per cent overall net rental growth
- The leasing surge is happening even as the broader debate continues over whether AI will ultimately create or destroy office-using jobs in the long run
A few years ago, London’s office market was wrestling with the aftermath of remote work adoption, tech sector hiring freezes, and rising vacancy rates in buildings that had been full before the pandemic. That conversation has shifted noticeably in 2026, and the reason is artificial intelligence.
It was reported on April 22, 2026, that AI companies actively seeking office space are giving London’s commercial property developers a meaningful boost to their near-term outlook. The data behind that reporting landed concretely through British Land’s latest trading update, which showed the property giant raising earnings guidance for both the current and next financial year on the back of what its CEO described as “accelerating demand from a new wave of AI and innovation-led occupiers.”
The biggest deal in recent weeks belongs to Anthropic, the maker of Claude. Confirmed that Anthropic agreed to lease 158,000 square feet at British Land and Royal London Asset Management’s One Triton Square development near Euston station, a space sized for roughly 800 people. That is a significant expansion from Anthropic’s existing London office at 107 Cheapside, which accommodates around 200 people. The One Triton Square deal effectively fills the 300,000 square foot building to near capacity.
Not far away, OpenAI has taken 89,000 square feet at Nan Fung’s Regent’s Quarter development in King’s Cross, having already designated London as its largest research hub outside the United States earlier this year. Add several other AI-linked leasing deals across the city, and the total reaches well over 375,000 square feet in a concentrated period, according to tracking of London office activity.
British Land is the clearest financial beneficiary in the listed property sector. The FTSE 100 company reported 12 percent like-for-like rental growth across its campus properties, including Regent’s Place, Broadgate, and the Triton Square developments, for the year ended March 2026. Overall, like-for-like net rental growth came in at 6 per cent. The company has now raised underlying earnings per share guidance to 28.9p for fiscal 2026 against previous guidance of 28.5p, and lifted its fiscal 2027 target to at least 30.5p from 30.2p. CEO Simon Carter said the market remains supply-constrained, meaning the demand pressure from AI tenants is not being absorbed by a flood of new buildings. The pipeline is limited and the competition for quality space is real.
London’s attraction for AI companies is not hard to explain. The coverage of OpenAI’s expansion pointed to the city’s deep pool of AI research talent fed by leading universities, established relationships with European regulators, and proximity to enterprise customers across EMEA. For companies building and scaling AI products that need to engage with governments and large institutions, London offers a combination of things that very few other cities can match.
The longer-term question hanging over all of this is one the property industry has not fully answered yet. AI is currently driving office leasing demand in London at a pace that is visibly moving landlord earnings. But AI is also, by design, a technology being built to automate work that currently requires office-based employees. Whether the net effect on office demand over a five-year horizon is positive or negative remains genuinely unclear.
For now, the landlords are not complaining. And neither are the developers with large London campuses sitting in a supply-constrained market with Anthropic on one side and OpenAI on the other.
