job search.

buildspace - blog

ai is changing office fit-out twice over, and both changes land on your hiring.

Back to in:SITE


For a while, the assumption was that AI would empty offices out. Fewer desk jobs, less space needed, less work for the people who build those spaces.

The leasing data says something close to the opposite. AI companies have become one of the biggest new sources of demand in the London office market, and the way they take space, along with the way that space now needs to be built, both point back to the same place: the teams doing the work.

Here's what's actually happening, and what it means if you're resourcing a fit-out business.

1. The occupiers: AI firms are taking real space, and taking it permanently

The numbers are no longer sentiment. According to Knight Frank, AI companies took just over 661,000 sq ft of London office space in the first half of 2026 alone, more than the roughly 500,000 sq ft they took across the whole of 2025. Since the start of last year, the sector has committed to close to 1.2 million sq ft in the capital. Knight Frank expects the 2026 total to land near 1 million sq ft: double 2025, and around four times the 2024 figure.

The headline deals give a sense of the scale. Anthropic has taken 158,000 sq ft at One Triton Square in Regent's Place. Databricks has committed to over 137,000 sq ft at Network W1 in Fitzrovia for its EMEA headquarters. OpenAI has taken close to 90,000 sq ft in King's Cross. The average AI deal in H1 was around 22,800 sq ft, and Knight Frank is tracking a further 610,000 sq ft of live requirements, with 240,000 sq ft under offer.

The detail that matters for fit-out: these companies are moving out of flexible and serviced space and into large, permanent headquarters. That is not a desk-rental decision. That is a CAT B project, and usually a demanding one, for a client with funding, a brand to express and a hiring plan to house.

The demand is clustered around King's Cross, Euston and Fitzrovia, where occupiers sit close to universities, research institutions and transport. And it's landing in a market with almost nothing spare: prime vacancy has fallen to around 0.3% in the City core and 0.8% in the West End core. With so little available, a number of AI occupiers are pre-letting, committing to buildings years ahead of completion. For contractors, that means pipeline visibility further out than usual.

2. The specification: AI is changing what a fit-out actually is

The second shift is quieter, and arguably bigger, because it doesn't only affect AI occupiers. It affects everyone's projects.

Turner & Townsend's research finds that AI is pushing occupiers towards higher-quality, more amenitised and more sustainable workspace. Businesses no longer want technology bolted on at the end. They want it integrated into the fabric of the building, so they're ready for whatever comes next. Amenities themselves are increasingly AI-driven, with climate control, lighting and space usage all digitally tracked and connected.

JLL's 2026 fit-out cost guide puts the same trend in numbers. Global fit-out costs are up between 2% and 6%, with AI-readiness now a baseline requirement to be budgeted from day one rather than a nice-to-have. Across EMEA, M&E services and Security, IT & AV works are the fastest-growing cost categories, and together now account for more than 40% of total fit-out spend on average.

Put plainly: a bigger share of the job is now services and technology. The work is more technical than it was three years ago.

3. Where this becomes a hiring problem

Follow those two threads and they meet in one place.

More projects, more technically complex projects, and, according to JLL, an acute shortage of exactly the people who deliver them. Skilled trades are the constraint, with electricians, M&E specialists and technology installation teams named as the tightest of all. Labour shortages are affecting a majority of markets globally, and JLL notes that AI infrastructure expansion, data centre construction in particular, is competing for the same materials and the same skilled trades.

So the fit-out firms best placed to win this wave of AI-driven work aren't necessarily the ones with the best pitch. They're the ones who can actually staff it: M&E leads, project managers who've delivered tech-heavy CAT B, and installation teams who won't be poached mid-programme.

That is the bottleneck. Not demand.

What we'd suggest

If you're a fit-out contractor looking at the next 18 months:

  • Treat resourcing as a bid strategy, not a back-office task. With pre-lets being signed years ahead, the pipeline is visible earlier than usual, which means you can hire ahead of it rather than scrambling behind it.
  • Weight your hiring towards M&E and technology. That's where the spend is growing and where the shortage bites hardest.
  • Know what you're paying. In a market this tight, the difference between winning a hire and losing one is often a benchmark you didn't have. (We've just refreshed our office fit-out salary benchmarks, so do ask us for a copy.)

The AI boom isn't an abstract macro story for this sector. It's showing up as signed leases, tighter specs and harder hires. The demand is coming. The question is who's resourced to catch it.

Build Space specialises in recruitment for the office fit-out and workplace sector. If you're planning your team for the projects ahead, we'd be glad to talk.

Sources

All figures as reported July 2026.

ready to work with us? let's talk!

subject:
  • I'm looking to hire
  • I'm looking for work
  • I'm interested in working for Build Space
  • I'm looking for information on IR35

  •