Half of London's businesses now say their workforce does not have all the skills it needs to meet their requirements in the age of artificial intelligence. As reported by the BBC, a Survation survey of more than 2,000 business leaders for BusinessLDN put that figure at 50 percent, down from 63 percent a year earlier. The share of firms reporting significant skills and capacity gaps reached 15 percent, up from 4 percent in 2025, and the highest the annual survey has recorded.
What the survey found
The headline is not that London has discovered AI. Three-quarters of the businesses surveyed said they were already using it in some form, and only 5 percent said they had no plans to. The headline is that adoption has run ahead of capability. Among firms already using the technology, 85 percent said it had changed the skills their workforce needs, with a greater emphasis on critical thinking, ethical reasoning, and decision-making rather than narrow technical ability.
The gap is concentrated in digital skills. Sixty percent of the firms reporting shortages said they lacked advanced digital skills, and 78 percent expect a significant need for them over the next two to five years, up from 66 percent last year and 56 percent in 2023. The direction of travel is clear, and it is accelerating.
Why the gap is widening, not closing
It would be easy to read this as a training problem. It is not, or at least not only. When a firm adopts AI tools without first deciding which processes should be automated, who is accountable for the output, and how decisions will be checked, the result is activity without control. People end up using tools they do not fully trust, on tasks that were never scoped, with no audit trail. That feels like a skills gap, and in part it is, but the deeper gap is one of governance and judgement.
This matters more for the lower-mid-market than for the largest firms. A FTSE 100 company can stand up an internal AI function and absorb the cost of getting it wrong. A business of twenty to two hundred people cannot. For that firm, the question is not how to train everyone, but where to start so that the first AI investments actually pay back.
What this means for a mid-market business
The firms that handle this well tend to do three things. They take an honest inventory of where their people actually spend time, and which of those tasks are repetitive and rules-based enough to automate safely. They put a small number of governed, auditable workflows in place, rather than rolling out a general-purpose tool and hoping. And they keep a person accountable for every output that leaves the building. None of this requires a large team or a long programme. It requires a clear assessment and a disciplined first step.
How we think about it
At Blash Advisory we treat this as a readiness question before a technology question. An AI readiness assessment looks at your processes, your data, and your governance, and identifies the handful of workflows where automation would remove real cost without creating new risk. From there we install those workflows on a fixed fee, with audit trails and clear ownership built in, so the capability sits inside your business rather than in a tool you rent. The point is not to use more AI. It is to use it where it pays, and to use it safely.
Where to start
If your team is feeling the skills gap that this survey describes, the first step is not a training budget. It is an honest look at which processes are ready to automate and which are not. If you would like clarity on where to start, book a consultation.
