Why your AI strategy needs a people strategy
De crisis in vaardigheden is geen technologisch, maar menselijk probleem. Om dit op te lossen, moeten we AI niet langer als een op zichzelf staande strategie beschouwen. Sarah Mason, Chief People Officer bij SThree, licht haar visie over dit thema toe.
(This blog continues in English)
Four in five STEM professionals say skills gaps are directly affecting their organisation’s revenue. That statistic alone tells us something fundamental has changed. Skills have moved from a second-order HR concern to a strategic business risk tied to growth, competitiveness and survival.
Yet when I look at how most businesses are approaching AI, I see a significant gap. They have AI strategies, yes. But too often, these are really just technology strategies – focused entirely on what the tech can do, without considering the people who need to work alongside it.
An AI strategy that doesn’t include a people strategy isn’t complete. It’s not the full picture. And this matters because we’re at an inflection point – and how we respond will determine which organisations lead the next decade and which fall behind.
Start with the outcome, not the technology
What are we trying to achieve through AI? That’s the question I always come back to. A lot of the time, we are looking for AI to deliver tasks. Businesses outcomes can be broken down into a series of tasks to be delivered. Historically, we’ve assumed people would do those tasks, but now there’s another option.
Most organisations start with the technology and ask what it can do. Another approach is to start with the business goal, map the tasks that deliver it, and then decide which tasks sit best with AI and which with humans – and crucially, how they work together.
This isn’t just an IT decision. Finance must weigh affordability, HR must understand capability gaps, and Operations must map workflows. A genuinely effective AI strategy is cross-functional and collaborative, not siloed in a technology team. It’s an augmented approach – where people and technology work in concert, not in competition.
Think about AI as another “destination” for work, much like outsourcing or offshoring. Framing it this way demystifies the conversation and makes it easier to have practical discussions about what sits where.
The obsolescence myth
Our STEM Workforce Report shows that a third of STEM professionals expect the skills they’ll need most in the next 12 months to be AI-related, and many believe traditional technical skills such as data analysis and coding are quickly becoming obsolete or redefined. It’s easy to see why anxiety is running high.
But I’d urge caution about the narrative of mass skills obsolescence. While the shift is accelerating, fears of wholesale replacement are often overstated. AI skills are emerging as essential enablers, certainly, but core technical skills remain foundational. It’s important to see how both people and AI work together; it’s focusing on the ‘and’ not the ‘or’.
Organisations should focus on blending AI literacy with traditional and existing expertise, not pitting one against the other. It’s less about substitution and more about integration.
AI skills are emerging as essential enablers, certainly, but core technical skills remain foundational. It’s important to see how both people and AI work together.
The trust paradox
One striking finding from our research surprised me: 68% of STEM professionals support organisations outsourcing skills and knowledge to AI agents. That suggests growing trust in AI as a productivity partner. Something to collaborate with, rather than fear. People are recognising the opportunity to be freed up for higher-value work – the complex, creative, strategic undertakings that humans do best – while AI can handle those repetitive, routine tasks.
At the same time, over-reliance on AI risks eroding our critical thinking. If we get AI to do our thinking for us, what does that mean for our ability to reason in the future? We must protect human judgement and ethics, all the more so as we embrace AI’s capabilities and scale what it can do.
Think tasks, not jobs
Rather than thinking of a job as a single unit, I find it helpful to acknowledge that jobs aren’t monolithic. They are bundles of tasks, and when AI arrives, some of these tasks will go away, and others won’t.
Take job specifications, something every HR professional knows intimately. Drafting them isn’t particularly enjoyable – AI can certainly help with that. But someone still needs to provide inputs, review the output, ensure the tone of voice aligns to the culture, and sign off. Even something as simple as a job spec still requires human judgement at multiple points.
The pattern repeats across roles. Some tasks will be automated while others won’t. What changes is the proportion and nature of the work. Instead of thinking about jobs disappearing, we should be thinking about jobs evolving, and supporting people to evolve with them.
When you frame it this way, the conversation becomes more productive.
Some tasks will be automated while others won’t. What changes is the proportion and nature of the work. Instead of thinking about jobs disappearing, we should be thinking about jobs evolving, and supporting people to evolve with them.
Instead of thinking about jobs disappearing, we should be thinking about jobs evolving, and supporting people to evolve with them.
The psychology of change
I often return to Paul Strebel’s research. His central finding was that leaders see change as opportunity, but many employees experience it as disruption. What’s particularly important is that in his research, middle managers – the people we typically expect to lead change – also fell into the “disruption” camp.
This matters enormously for how we approach AI transformation. If your workforce views change as disruptive, know that that’s entirely normal and reasonable. People have invested years developing expertise, accumulating certifications and forming identities around what they know. Telling them to let go of that is asking a lot.
The healthiest framing isn’t to dismiss those concerns but to acknowledge the disruption while offering a path forward, positioning this as career evolution rather than a threat.
Practically, that means being transparent about what’s changing, investing in reskilling and redeployment, enabling two-way communication rather than top-down announcements, and strengthening psychological safety so people feel heard.
Over half of organisations have made or are planning redundancies due to AI, which has created an environment where maintaining trust requires careful management, including upskilling and redeploying where possible. Communicate your how and why well – and give people the space to respond, question and contribute.
A positive outlook
When I look at how organisations are responding, I see genuine reasons for optimism.
Some are already investing seriously in skills-based workforce planning. L’Oréal, for instance, has spent several years developing a skills-based approach across their complex global footprint. Their CHRO spoke recently about the journey – it’s impressive and demonstrates that even at scale, this move is possible.
Fundamentally, I remain confident that certain capabilities will remain distinctly human: ethics, judgement, creativity, critical thinking – and the ability to check AI’s work and ensure it’s accurate and appropriate. Rather than being peripheral skills, they are central to the necessity of human contribution.
AI is undoubtedly powerful, but its value depends entirely on the human capability to deploy it well. The organisations that invest in developing those skills – alongside the technology – will find themselves with a workforce that’s not just surviving this transformation, but leading it.
AI is undoubtedly powerful, but its value depends entirely on the human capability to deploy it well.
Your company is your people
I often hear business leaders talk about getting “the company” ready for AI, as if that’s somehow separate from getting employees ready. But companies rest squarely on the shoulders of a group of people. If the employees aren’t ready, then neither is the company.
Building an AI-ready culture means focusing on skills, mindset and governance together – and ensuring your people strategy is made explicit within your AI strategy.
Adaptability will define the next decade. The professionals who thrive won’t necessarily be those with the most technical certifications, but those who can evolve continuously, integrate new tools and navigate uncertainty with resilience.
Adaptability will define the next decade. The professionals who thrive won’t necessarily be those with the most technical certifications, but those who can evolve continuously, integrate new tools and navigate uncertainty with resilience.
The future of STEM work is being formed right now. The only question is whether your organisation will help shape it – or be shaped by it.
About the research
SThree partnered with YouGov to survey 5,391 STEM professionals across the UK, US, Germany, the Netherlands, UAE and Japan between July and August 2025. The research covers professionals working in IT/technology, engineering and life sciences across organisations of all sizes.
Source: SThree
For more information about SThree, click here.
Lees meer:
