The demand signal is loud. Over the last year, demand for AI skills climbed nearly 200 percent, with London alone accounting for 80 percent of AI-related job postings and nearly two-thirds of all technology vacancies in the UK, according to Accenture data reported by The Register. That concentration tells you something: the talent you want for a controls role sitting inside an AI transformation programme is being chased hard, often by employers who move faster than you.
The wider market is more nuanced than the headline. Against a subdued 2025 economic backdrop, businesses took a cautious approach to hiring, yet there were clear demand hotspots for specialist skills including AI, data, enterprise applications and cyber security, as set out in the tech recruitment outlook for 2026 from Computer Weekly. Controls work sits right in the middle of those hotspots, because governing an AI system draws on data, risk and engineering skills at once.
So the question is not whether demand exists. It is how you compete for a narrow pool of genuinely capable people without over-indexing on a confident interview or a familiar CV. That is where instinct lets teams down and evidence earns its place.
The volume is set to grow. Across AI-related occupations, projections show a net change ranging from roughly 696,000 jobs in one scenario to 1,072,000 in another, with replacement demand rising from 3,077,000 to 3,193,000 jobs, according to GOV.UK labour market and skills projections. A controls or governance specialist draws on several of these occupations at once: data handling, risk, model oversight and the engineering discipline to put guardrails in place.
The routes into these roles are widening too. Apprenticeships have risen from 3 percent of AI hires in 2020 to 19 percent in 2025, according to the AI Labour Market Survey 2025 from GOV.UK. That matters for a controls hire, because it tells you that strong candidates increasingly arrive without the degree-and-big-brand path you might instinctively screen for. Judge them on demonstrated capability instead.
Supply is being built deliberately. During London Tech Week 2025, the UK government announced a partnership to upskill 7.5 million UK workers in essential AI skills by 2030, equivalent to around 20 percent of the UK workforce, as reported by techUK. More trained people is good news, but it also means your assessment has to be sharp enough to tell real controls competence from a freshly minted certificate.
Start by writing down what good looks like before you meet anyone. Define the controls the role must own, the systems they govern and the decisions they will make under pressure. When your scorecard exists before the first conversation, you stop grading charisma and start grading fit against the job.
Then test the work, not the talk. Use a short, realistic task: review a model's monitoring gaps, draft a control for a risk scenario, or walk through how they would catch drift. Ask the same structured questions of every person so you compare like with like. Given that apprenticeships now make up a far larger share of AI hires than five years ago, keep the entry criteria about proven skill rather than pedigree.
Finally, move at the speed of the market. With AI demand up nearly 200 percent in a year, a slow process loses good people to faster employers The Register (Accenture data). Evidence and speed are not in tension: a tight, consistent scorecard lets you decide quickly and defend the decision afterwards.
This is where we help. We search 15 million candidates, rank a shortlist against the capabilities you set, and surface the evidence behind each match so you decide on what someone can do, not how they came across. You stay in control of the brief and the bar; we do the heavy lifting of finding and ranking the people who meet it.
Our recruitment agent manages recruitment end to end for 8% on a successful hire, with no monthly fee and no upfront cost. If your next controls hire is too important to leave to instinct, tell us what good looks like and let us bring you the evidence.
