AI now handles tasks that used to consume entire afternoons: drafting emails, analysing spreadsheets, summarising reports. As routine work shifts to machines, the skills that make employees valuable are changing too. The answer isn't technical expertise, according to new guidance from HR technology provider HiBob. It's the human judgement around AI's output.
When technology can process information and produce content in seconds, employees earn their keep by questioning results, shaping them, communicating them and knowing when not to trust them. That means AI readiness is about more than giving people access to tools. Organisations need to develop the human capabilities that help people get better results from the technology.
Giving employees access to AI and teaching them how to use it is only part of the picture. Workforce transformation still lags behind AI investment, Toby Hough of HiBob said in a piece for HR publication Business Reporter. HR leaders need to keep five skills in their workforce as stretch AI adoption continues.
Critical thinking: don't outsource judgement
AI can provide answers in seconds. And when the output is immediate, well-written and appears evidence-based, it's easy to give it more authority than it deserves. But a convincing answer isn't necessarily a good one.
Employees need to interrogate what AI gives them, question its assumptions and decide whether its conclusions make sense. The risk of not doing this is that people become accustomed to accepting answers rather than forming their own judgement.
Employers need to make that judgement a deliberate part of AI training. Give employees AI-generated recommendations and ask them to identify what's missing, question the evidence and explain where they disagree. Managers can reinforce that by asking teams to show the reasoning behind their decisions, not just the outputs.
Scoping: know what problem you're solving
Prompt engineering has long been a staple in AI training programmes. But the ideal prompt will change as the technology does, Hough argues. A more durable skill is knowing how to define the problem you're trying to solve.
Business challenges are rarely delivered as clean instructions. Employees need to take ambiguity and create clarity, defining the outcome they're after and separating symptoms from underlying problems.
Organisations can develop this by making problem definition a core part of AI training. Give employees ambiguous business challenges and ask them to scope the problem before touching the technology. Managers can ask, "What are we trying to change?" or "How will we know if we've succeeded?"
Writing: invest in editors, not generators
AI has made producing written content easier than ever. The valuable skill is no longer generating the first draft but knowing what deserves to survive, Hough said.
Employees need editorial judgement: the ability to recognise what's important, remove what isn't, adapt information to the right audience and spot text that sounds polished but says very little. As audiences get better at recognising AI-generated content, the ability to decide what's actually worth saying carries more weight.
Writing training should evolve alongside that shift. Programmes should give employees AI-generated drafts and ask them to sharpen them or rewrite the same information for different audiences, rather than focus on raw generation. If your organisation is early in this journey, a structured AI course for HR managers can help you think through the training approach before you build a programme.
Storytelling: turn information into influence
More insight doesn't mean better decisions, Hough says. Employees still need to establish what matters, why it matters and what should happen next. That requires more than presenting what AI produced; employees need to turn information into a clear, compelling rundown.
Best developed, the skill comes through practice. Give employees opportunities to present AI-assisted analysis, make a recommendation, defend their reasoning and respond to questions in real time. Teams can also practise delivering the same insight to different audiences.
Emotional intelligence: practise the one skill AI can't
AI can learn many skills, but emotional intelligence currently isn't one of them. Interpersonal relationships, difficult feedback, and earning trust remain fundamentally human - and that becomesmore important during technological change. As AI adoption accelerates, employees will want questions about its impact on their roles answered. The quality of those responses affects how they engage with change.
"At no point do HR practitioners need to be more skilled in communication than when managing the transition to AI," its effect on roles, expectations and futures are top of mind, working teams need space for genuine human discussion when topics are complex or sensitive.
Why this matters for HR leads
Workforce transformation still lags behind AI investment. HR leaders can close that gap less by pouring time into skills that get automated anyway. Use the five skills above to re-align your AI training to what machines can't take away: critical thinking, problem definition, editorial control, storytelling and emotional intelligence.
HiBob's guidance is a practical framework. Take the exercises it suggests, apply a test on an ambiguous problem rather than a prompt engineering session, and put an AI draft on the table with a request to cut it back. That's how AI skills and human skills don't become the same conversation about tech to talk about - people. HR leaders can build a workforce that works with AI without losing the human edge.
Your membership also unlocks:
