A new Lloyds Banking Group survey has put some welcome, if complicated, numbers on the AI-and-jobs debate. Fifty-four percent of UK businesses say AI has already created new jobs within their organisation, 61% are now using the technology in some form, and 58% plan to invest further in AI skills over the coming year. Large enterprises are leading the way, with 79% adoption among firms turning over more than £10m, against a much lower rate further down the size scale. But the same survey found 31% of businesses admit their workforce doesn’t have the AI skills it needs — a material shortfall even as investment climbs.

For SMEs, this is a more useful story than either the “AI is taking all the jobs” headlines or the “AI is magic productivity” hype. The real pattern emerging is that AI adoption is creating new kinds of roles and responsibilities faster than most businesses, particularly smaller ones, can train people to fill them. That gap, not the technology itself, is where the actual business risk and opportunity sit right now.

What firms are actually doing about it

Among businesses already responding to the skills gap, 43% are introducing AI skills training, 32% are expanding training they already run, 24% are prioritising AI skills when hiring, and 21% have created new AI-specific roles outright. None of that requires enterprise-scale budgets. A firm with ten staff can decide that one person owns “how we use AI tools properly” as part of their role, and that’s a meaningfully different position than having nobody responsible for it at all.

Closing the gap without a hiring spree

Don’t wait for a dedicated hire to start building AI skills. Training existing staff on the tools you already have, Copilot, ChatGPT, or your CRM’s built-in AI features, closes more of the gap faster than an open vacancy that takes months to fill. BuildApps works specifically with UK SMEs on practical AI adoption and the training that goes with it, rather than treating AI as a bolt-on nobody was trained to use.

Turn over £1m and feel behind? You’re the norm, not the exception. The survey found competitive-disadvantage concern about AI is far higher among larger firms (73%) than smaller ones (54%). That’s not smaller businesses being complacent, it’s that the case for AI adoption genuinely looks different at different scales. Adopt what solves a real problem you have, not what a case study from a much bigger company suggests you should.

If a new AI-related role or responsibility is worth creating, define what it actually covers before advertising it. A vague “AI lead” job spec attracts the wrong candidates and leaves existing staff unclear on what changed. Decide the specific tasks, tools, and decisions the role owns first.

Where the skills gap actually bites

The 31% of firms admitting a skills shortfall aren’t necessarily the ones without any AI tools switched on, that’s often the more dangerous middle ground: AI adoption has happened, but nobody has been given the time or training to use it properly. That looks like staff quietly avoiding a tool they were never shown how to use well, or worse, using it in ways that create risk, pasting sensitive client data into a public chatbot, trusting AI-generated output without checking it, because nobody set out what good practice looks like. The gap Lloyds is measuring isn’t really about hiring; it’s about the difference between a business that rolled out a tool and one that rolled out a tool with a plan for how people actually use it day to day.

The takeaway

The Lloyds numbers say AI is already reshaping UK hiring, quietly and unevenly, and that the businesses feeling the benefit are the ones that paired adoption with training. If your business has added AI tools this year without a plan for who’s actually skilled to use them properly, that’s the gap worth closing before the next headline about AI and jobs makes it feel more urgent than it needs to.