For some businesses, AI implementation follows a familiar pattern: give staff the tools and hope the rest will follow.
According to leaders in HR, AI investment and technology law who spoke at a leadership discussion in Leeds, that approach tends to leave teams feeling like they have more to do, not less.
The panel, titled “The Leadership Challenge of AI: People, Confidence and Change,” took place at Climb26, the West Yorkshire business festival, in early July. Moderated by broadcaster Tanya Arnold, the 40-minute session examined how growing businesses introduce AI in the workplace without losing staff trust.
Clair Flynn, Employment Hero’s Manager for People Partnering, emphasised the assistive role of AI over its capacity to replace judgement. Speaking to the presumption that adopting these tools should mean less hands-on involvement from staff, she stressed that the real advantage is speed, not automation: “It enhances what you can do in your day, your role, and becomes your assistant. But it still needs human handling… it still requires real professional expertise.”
That handling, she says, is where most businesses fall short: “The worst thing that any business or organisation can do is just start dropping AI tools around without any sort of support or any sort of training.”
In terms of mitigating the potential errors AI can make, Clair stressed that fact-checking remains essential. “I think people realise that now, and I see a lot more people starting to become more cautious and reviewing outputs in much more detail.”
Troy Wood, AI Investor at EHE Ventures, applies the same caution to his own use of AI when sifting through market data: “It gives me a really good base. I’m still double checking that output to make sure it’s fine… I’m not asking it for opinions. I’m just getting more articulate summaries that allow me to make a quicker decision.”
AI Skills Now Rank Ahead of Experience in Hiring – 23% of UK Workers Feel Unprepared
New Employment Hero research shows UK employers are already prioritising AI skills when hiring, but speakers at Climb26 argued the technology will only earn trust once leaders replace guesswork with honest guidance. Nearly two thirds (64%) of UK employers now say AI in the workplace has changed what they look for in candidates, with AI skills entering the top five attributes they screen for, ahead of prior experience.
That shift is moving fastest at entry level: 37% of UK workers say junior roles increasingly specify AI knowledge, though almost a quarter (23%) don’t feel their own skills are strong enough to compete. For businesses following Clair’s advice, the gap between what employers now expect and what workers feel able to offer is exactly what structured training is meant to close.
How AI Upskilling Works When It’s Built Around Roles, Not Tools
One recommended approach, from a people perspective, is being deliberate about who leads the way rather than training everyone at once: identify who adapts to new tools fastest, and build outward from there. “If you can find those individuals and figure out how to work with them, and help them get the best out of how they can use those tools to enhance overall performance, I think that’s a good point for us as well,” Clair says.
AI’s flexibility to fit almost any business process cuts both ways, Clair says: that same adaptability means it takes real knowledge to make it work well, not just access to the tools. “It’s a system that you can build your own bespoke operating system around, but not without the knowledge of how to do that,” she says.
Another recommended approach is reconsidering processes like onboarding altogether: phasing tools in gradually rather than expecting fluency from day one, so staff build confidence over several weeks as they move from getting comfortable with the basics through to more hands-on practice. Clair points to how that’s worked in practice: “We completely rebuilt our onboarding process,” she says. “Our team goes from having an idea about how we would use it, to really fully understanding it and starting to build out a personalised capability plan… We ran a six-week accelerator through the business… in weeks one and two you’re learning the fundamentals around AI and the tools we use, and by weeks four, five and six you’re building agents, creating prototypes, and you’re starting to get that confidence and ability built in.”
Not everyone moves through that process at the same speed, and Clair says few training plans account for the difference. Some staff will teach themselves by experimenting; others, in more customer-facing roles, need a different route in. “People learn at different speeds,” she says. “It’s important to understand it’s different for every role as well… You’ve got people who can naturally spend a lot more time just playing around and making and breaking things. Whereas when you look at a more customer-facing team, their adoption journey is completely different. You have to factor that into your training expectations.”
Training begins with a question rather than a tool, Clair says: “What outcome are you expecting to see on the other side of this? What do you want it to achieve? Where are you looking for it to benefit you?” Skip that step, she suggests, and businesses end up building fluency with the software rather than confidence in using it.
Only 40% of Leaders Feel Confident Designing AI Reskilling Pathways, CIPD Finds
Confidence gaps like this aren’t unique to Climb26’s audience. CIPD research published this year, based on a YouGov survey of more than 1,300 senior leaders and HR professionals, found many leaders feel confident setting direction on AI but less confident with the practical follow-through: only around a third feel able to estimate the future workforce needs it creates, 46% feel confident leading job redesign and 40% feel confident designing reskilling pathways. For most, that gap reflects how new this all is rather than any shortfall in ability.
Dahlia Stroud, Director at SBWD Group, says the priority is to “build the business” and “build the right talent” – not the tools themselves: “We’ve got to do the first step, which is to prioritise the business.” Efficiency isn’t the only goal, she adds: “We want the tools to be fun.”
The adoption of AI tools is also a financial opportunity for many businesses. The UK government’s AI Opportunities Action Plan estimates AI adoption could grow the UK economy by an additional £400 billion by 2030. Stressing the history of automation freeing up money for growth in business, Troy argues businesses have been chasing efficiency “in every aspect of life” long before AI arrived. “We’ve done that since the industrial revolution – automation and robotics have been a huge part of it,” he says. “Take the IKEA example, or Amazon. From a leadership perspective, that’s a different way of looking at it – you can get a fundamental, massive cost reduction, and that allows you to put more resources into growing the business as it progresses.”
Clair, too, pushed back on the idea that automation removes people from these processes entirely: ” there’s still opportunityto have that human in the loop, which absolutelyhas to exist. Our people and our relationship builders – they’re not going anywhere, they’re still there. Even with AI tools and automated systems, we still need our teams; we haven’t lost that from our organisation at all.”
Dahlia, from SBWD Group, drew a similar parallel with the advent of online shopping: “We’ve seen this manifest elsewhere so many times. I think about retail as an example when everyone started shopping online and everyone worried ‘will it be the death of the high street.’ But then suddenly there was a renaissance of retail in a different format through more experiential things, finding community and making life easier.”
Why the Fix to Unauthorised AI Use at Work Is Clarity, Not Crackdown
Working around unclear rules isn’t limited to older generations, either. The same Employment Hero research – the Focaldata poll of more than 3,500 UK employers and workers cited earlier – found four in ten Gen Z workers already use AI without their employer’s knowledge. Clair argues that’s a governance failure rather than a discipline problem. The employees quietly using unauthorised tools, in her experience, tend to be “some of the most ambitious and most curious people that you’ve got in your business,” not the ones cutting corners – and most avoid asking permission because they assume they can’t, don’t know who to ask, or are too scared to find out.
The fix, in her view, is clarity rather than restriction: set out clear approved tools and accounts “so it’s incredibly clear from the start: this is what we want you to use, this is what we want you to start experimenting and playing with, and this is what we expect to come out of it.” What doesn’t work, she says, is punishing people for using AI at all: ” the worst thing you can do is go in and take a real hard stance on anybody who’s trying to use these tools – because most of the time, they’re just curious.”
She frames the wider anxiety as something leaders can plan for rather than just absorb: “As much as you want to address that fear and that anxiety around roles changing, change is inevitable. A lot of this is an intentional change management exercise… it’s an education piece, overall. You’ve got to set yourself up as best as you can by providing a clear framework and strong training support.”
For SMEs weighing up their next step on AI, the panel’s advice was less about which tool to buy and more about the culture and training wrapped around it. From their point of view, if you give your workforce structured time to learn and honesty instead of guesswork, the confidence to use these tools safely tends to follow.
