TL;DR

Ola Obadara runs capital projects for the City of London Corporation, between 40 and 50 of them at any moment, and describes herself as an AI optimist. Her actual use of it stops at Copilot for research. Her most useful contribution is an argument against the industry’s favourite robot.

The dishwasher argument

Having attended a conference on humanoid machines in construction, Obadara came away suspecting the sector is asking the wrong question. Human-shaped robots will do a decent job of scanning, programming and snagging, she accepts. But nobody answered the problem of washing up by building a mechanical person to stand at the sink; somebody designed a different process instead. Reproducing site labour in humanoid form, on her reading, skips the harder and more valuable work of rethinking how you build around what the technology can genuinely do.

Where clients sit

Her wider case concerns position in the hierarchy. Clients set what success means and align the outcomes, which is why she believes collaboration on research and adoption has to begin there rather than with contractors, and why she names shorter programmes and lower costs as the prize. She is candid that the Corporation has not arrived: keeping pace is the ambition, and how best to put any of it to work remains unresolved. Her own use is modest, asking Copilot when a particular type of building was last done or what it cost, instead of losing hours to searching.

The speculation worth watching is procurement. Drawings might be scanned to produce a bill of quantities; bids might be assessed by machine as well as drafted by one. Asked whether AI-written tenders raise an ethical problem, Obadara says no, reasoning that most people are still using these systems to look things up. That answer will date quickly.

Looking forward

Construction seldom features in AI coverage, which makes a named public client worth hearing. The Corporation buys through committee — 125 elected members, reports going up at least weekly — and is already absorbing hold-ups from the Building Safety Regulator alongside inflation and shortages of both labour and materials. That governance load, not the technology, is what actually paces adoption across the public estate, and it is precisely the constraint vendor research into UK public sector AI keeps declining to name. Obadara takes over as CIOB president next year, which will give the argument a platform.