Key Takeaways
Driving Office Demand
Signing Larger Office Leases
Clustering Around the Knowledge Quarter
Accelerating the Flight to Quality
Forcing Landlords to Invest in Retrofitting
Prioritising London Over Other Global Hubs
The Outlook for London's Office Market
AI companies are reshaping the London office market by signing some of the city’s largest office leases, clustering around a handful of key locations, and pushing landlords to reconsider where and how they invest. These findings are based on our own enquiries, transactional data, surveys, and broader market research.
Over the last decade, we’ve seen AI businesses grow from a handful of specialist tenants into one of the market’s most active industries, influencing rents, building standards, and competition for the best office space.
Our report discusses the drivers of demand, how landlords are responding, and what it means for London's office market going forward.
Key Takeaways
Here is a summary of the key findings from our research, which we explore further in the report.
- AI now makes up over 30% of the office space rented by tech companies in London, up from just 4% a decade ago, with AI leasing activity more than ten times higher than last year.
- A major landlord achieved profits as high as £294m, partly due to Anthropic signing a sixth lease with them, reflecting the dominance of AI in the market.
- Grade A office rents are hitting record highs, while Grade B rents have fallen by nearly 20% over the past year.
- Over $310bn has been invested in London tech since 2023, and the city now has more medium to large AI businesses than San Francisco.
- Cumulative AI take-up is expected to reach around 4m sq. ft. over the coming years, nearly triple where it stands today.
Driving Office Demand
Our market research indicates that the wider London office market is currently experiencing lower demand, with overall take-up across Central London running below its 10-year average and last year’s levels. In comparison, AI is the one industry bucking this trend, with enough leasing activity to shift the market’s headline numbers on its own.
Ten years ago, AI-based occupiers accounted for just 4% of London’s tech office take-up. Today, that figure has passed 30%, and from our enquiry figures, we can see that both quarterly and annual growth are accelerating rather than levelling off. Leasing volumes have also grown more than tenfold in a year, and we’ve noticed that take-up is on track to double again in 2026.
Based on companies we’ve found office space for, the average AI office lease in London is currently around 23,000 sq. ft., suggesting that deals at this scale are driving the numbers, rather than just a handful of large headline leases.
Our transactional data highlights that AI take-up of flexible office space in London is especially high, with around three-quarters of AI startups renting flex and serviced offices, rather than traditional leased space. A provider we work with tells us that close to a fifth of their flex space tenants are now AI companies, and many others we work with have noticed a similar trend.
Overall, this illustrates how AI has moved from a contributor to London office demand to the market’s primary driver.
Signing Larger Office Leases
We are starting to receive more enquiries from AI companies seeking larger workspace and longer terms in London, highlighting how they are increasingly renting office space as established, permanent occupiers do, rather than behaving like higher-risk, short-term tenants.
This is reflected in Anthropic’s leasing activity, with its latest deal at around 158,000 sq. ft. at One Triton Square in Euston. However, this is not a one-off, and it is their sixth lease with the same landlord as they have scaled up. This shows that AI businesses are reliable, long-term occupiers worth investing in.
OpenAI illustrates a similar pattern, having moved from flex space into a dedicated building of around 90,000 to 100,000 sq. ft. We have found that this graduation from flexible to traditional office space is becoming a template for how rapidly growing AI companies are expanding in London.
This behaviour suggests AI’s impact on the market is structural, rather than a short-term spike of one-off deals. 
Clustering Around the Knowledge Quarter
From our enquiries, we can see that AI companies mainly search for office space in the Knowledge Quarter. This area is centred around the academic, cultural, and research institutions in King’s Cross, Euston and Bloomsbury.
We found that around half of all AI office deals signed in Q2 2026 were in the King’s Cross and Euston area. This reflects its transformation into one of London’s most sought-after office districts, largely driven by AI companies choosing to base themselves near like-minded organisations and research talent from UCL, The Alan Turing Institute, and Google DeepMind.
However, Microsoft’s 100,000 sq. ft. Soho lease shows this clustering is not absolute and that AI-driven demand is starting to shape office markets beyond its core hub.
We are noticing that as office space to rent in the Knowledge Quarter becomes scarcer, demand is increasing in other areas of Central London. Clients we have worked with often start by looking in King’s Cross and then consider other areas, illustrating how AI is driving competition for space citywide.
Accelerating the Flight to Quality
Our research demonstrates that AI is also accelerating a wider shift in what tenants expect from their workspaces altogether, a trend known as the flight to quality. We’re increasingly receiving requests from AI-led businesses for advanced power capacity, connectivity, and digital infrastructure, leading them to choose premium, refurbished office buildings over older stock.
This is widening the gap between secondary and Grade A office space, something we’d expect to narrow over time, but limited supply is preventing this. Based on our understanding of the market and current demand, new-build construction would only cover about 1.3 years of typical take-up.
As a result, our research shows that businesses are chasing the same Grade A stock, pushing rents to record highs. This is based on our own enquiry patterns, as we have started receiving enquiries from different AI firms, alongside other law and finance-sector tenants, requesting to view the same limited pool of premium space.
At the same time, our data shows Grade B buildings are seeing rents fall by almost 20%. This is largely due to weaker network infrastructure, digital resilience, and green credentials. We expect this divide to widen even further as the 2031 deadline for minimum energy performance standards approaches.
Forcing Landlords to Invest in Retrofitting
In response to the rising demand for higher-quality space from AI companies, many of the landlords we work with are retrofitting offices by upgrading an existing building’s power capacity, connectivity, and energy efficiency, rather than starting from new.
This highlights how landlords are rethinking how and where they invest by matching what AI companies typically look for in office space - proof that the market is being reshaped, not just tenant demand and behaviour within it.
The payoff for landlords who retrofit early is already clear. Having spoken to the landlord at One Triton Square, we know that since being upgraded and securing Anthropic as a tenant, the building is now 78% let, with a further 16% under offer. As a result, we expect this to push other landlords to retrofit earlier to stay competitive for AI tenants.
Prioritising London Over Other Global Hubs
Many of the international AI companies we get enquiries from have told us they are choosing London over other global hubs for its talent, funding, and research ecosystem.
According to our research, over $310bn has been invested in London tech since 2023, with more medium-to-large AI companies than the San Francisco Bay Area (Silicon Valley), often considered the world’s leading AI hub. London has also reclaimed the top spot in Europe for tech and AI investment in 2026, overtaking Paris.
That said, both Paris and Munich remain significant rivals to London. Mistral’s Paris headquarters is larger than Anthropic's, and Munich's tech take-up as a share of its office market is also higher than London’s.
However, based on our market research, London’s total AI office leasing is below 2m sq. ft., compared with around 10.5m sq. ft. in San Francisco alone. Put simply, it is strengthening its status as Europe’s clear leader, powered by the same capital and talent driving AI’s growth in the city.
The Outlook for London’s Office Market
If AI office demand continues to grow at its current pace, our data suggests that in the coming years, cumulative AI take-up will reach around 4m sq. ft., almost three times its current level.
However, a recent survey we conducted indicates that the same firms leasing record amounts of office space are also cutting jobs elsewhere, using automation to replace the work people used to do. This highlights how AI is reshaping how companies lease office space and how they operate.
We’ve found that this surge in AI-driven leasing activity has also been compared to the dot-com bubble, and after speaking to providers and AI clients, they have told us that they wonder whether it will slow down just as quickly.
Despite this, our analysis shows that the area’s skilled workforce, world-class research institutions, and strong investment that first attracted AI businesses should help London withstand any downturn.
Over the past decade, we’ve seen AI evolve from a niche office occupier to one of the biggest drivers of change in the London office space market, influencing tenant expectations, where companies rent and how much office space costs.
For businesses seeking office space to rent today, the workspace and location you choose matter more as the gap between premium and secondary space widens. Grade A buildings are thriving, while Grade B stock struggles, highlighting how the market is being reshaped unevenly. If AI-driven demand keeps growing as expected, London’s office market will continue to shift alongside it.