Building an AI-ready team has almost nothing to do with hiring people who know how to build AI. IKEA turned 8,500 call centre workers into interior design consultants and generated an extra $1.4 billion in revenue. They did not hire a single data scientist to do it. They invested in the people they already had. That is the playbook.
There is a version of this conversation that goes something like: hire a machine learning engineer, buy some software licences, call yourself AI-first. A lot of companies are doing exactly that and wondering why nothing has changed.
The real picture is more interesting. IDC found that over 90% of global enterprises are heading for critical AI skills shortages by 2026, with $5.5 trillion in lost market performance on the line.[1] At the same time, 94% of CEOs say AI is their most in-demand skill, but only 35% feel their organisations have actually prepared people for it.[1] That gap between wanting AI-ready teams and actually building them is where most companies are stuck right now. And it is entirely fixable.
The number of jobs explicitly requiring AI fluency has grown sevenfold in just two years, from roughly one million roles in 2023 to around seven million in 2025.[2] You cannot hire your way to that scale. You have to grow it.
If you could only hire for one quality in 2026, make it this: the ability to learn fast and enjoy doing it. The AI landscape shifts quickly enough that what someone knows today will look different in eighteen months. What does not change is their appetite for figuring things out.
In practice, curiosity shows up before someone even gets to the interview. Did they pick up new tools on their own? Do they have opinions about what they are currently exploring? Do they ask better questions than the ones you prepared answers for? These are not soft signals. The World Economic Forum's Future of Jobs Report 2025 found that curiosity, creative thinking, and resilience are expected to grow in importance right alongside technical AI skills over the next five years.[3] The employers who have actually deployed AI at scale are telling you: hire the learners.
A tip for interviews: ask candidates about something they taught themselves in the last six months. Not something from a training programme. Something they were simply curious enough to go and learn. The answer tells you more than a CV ever will.
You no longer need a dedicated AI team to build an AI-capable organisation. What you need is a baseline level of AI literacy spread across every function, from marketing to finance to operations to HR. The person who used to wait for the data team to run a report should now be able to ask an AI tool the question themselves, check the answer critically, and act on it.
This is already mainstream. Job postings requiring prompt engineering skills more than doubled between 2024 and 2025, even for non-technical IT roles.[4] And 71% of business leaders say they would choose a less experienced candidate with strong AI skills over a more experienced one without them.[5] That is a significant shift in what the market is actually asking for.
AI literacy does not mean knowing how to train a model. It means knowing how to get useful results from AI tools, spotting when those results are off, and being able to work with AI outputs rather than just around them. It is a skill that can be taught. Which is good news, because you probably already have a team full of people who are ready to learn it.
Here is something worth saying plainly: AI tools are brilliant, fast, and occasionally completely wrong. They do not know what they do not know. They will give you a confident, well-formatted answer that is factually backwards, and they will do it without blinking. The most important quality in an AI-ready team member is not their ability to use the tools. It is their ability to interrogate the output.
This matters enormously in high-stakes situations. CVS Health uses automated auditing to keep its pharmacy AI consistently compliant with FDA guidelines, but the judgement calls that matter most still require people who know what good looks like.[6] The same is true in legal, in finance, in any domain where a confident wrong answer has real consequences.
One of the best interview techniques right now: bring in a real AI output, something plausible but with a subtle error or bias baked in, and ask the candidate to review it. Watch what they catch. Watch what they miss. Watch whether they trust it by default or approach it with healthy scepticism. That thirty-minute exercise will tell you more than six rounds of technical interviews.
Most companies do not need more coders. They need people who can work with data, guide AI tools, and explain results in plain language.
Workforce Institute, January 2026
Here is something the AI skills conversation gets wrong surprisingly often: deep domain expertise is becoming more valuable as AI gets better, not less. Generic AI ability is increasingly common. What is genuinely rare is someone who knows your industry inside out and can direct AI tools towards the problems that actually matter in your specific context.
Kelly's research found that two thirds of managers say their recent hires' biggest gap was not technical skills, it was industry experience.[7] More than 75% of AI job listings specifically ask for domain experts with specialised knowledge rather than generalists.[8]
Your experienced people who understand your customers, your processes, and your quirks are not at risk of being replaced by AI. They are your biggest AI opportunity. Equip them with the tools and the confidence to use them, and you have something genuinely hard to replicate.
In 2021, IKEA's AI chatbot Billie started handling 47% of their routine customer enquiries. That freed up 8,500 customer service employees from answering the same questions about delivery times and stock availability. IKEA faced a choice that most companies get wrong: cut the headcount, or invest in it.
They chose to invest. Those 8,500 people were reskilled as interior design advisors, a higher-value, more human role that required the warmth, contextual judgement, and product knowledge that AI simply cannot replicate. The result was an additional $1.4 billion in revenue.[10]
IKEA also launched a dedicated AI literacy programme for around 3,000 employees and 500 leaders, with modules covering AI fundamentals, generative AI, and responsible AI use.[11] The programme was later recognised by the European Commission as a best practice model. The lesson is not complicated: treat your people as the asset, not the problem.
Walmart is doing something genuinely remarkable. In 2026, the company is offering free AI training through Google's AI Professional Certification to all 1.6 million of its US and Canadian employees, as part of a commitment approaching $1 billion in workforce development.[12] Their CEO has been explicit: AI will not be used to cut headcount. It will be used to make their people better at their jobs.
Amazon has invested more than $1.2 billion in free skills training since 2019. Over 700,000 employees worldwide have completed programmes in cloud computing, machine learning, and AI fundamentals through AWS Training and Amazon's Machine Learning University.[13]
Microsoft launched its Elevate initiative in mid-2025, a $4 billion programme working with LinkedIn Learning, GitHub, and Code.org to credential 20 million people in AI over two years.[13]
You do not need Walmart's budget to apply Walmart's logic. Smaller businesses that build systematic AI literacy programmes for their existing teams consistently outperform those waiting for the perfectly credentialled external hire who, in many cases, simply does not exist yet.
As AI gets embedded into more consequential decisions, the ability to think carefully about what it is doing and who it might be affecting becomes a genuine competitive advantage. Compliance is the floor, not the ceiling. The organisations getting this right care about what their AI does to the people on the other end of it.
Look for people who ask uncomfortable questions. Who notice when a model might be producing biased results. Who care about how an automated decision lands for a real customer or colleague on the other end. IKEA's AI training includes dedicated modules on responsible AI for exactly this reason: ethics is not separate from capability, it is part of it.[11]
The World Economic Forum projects that ethical reasoning and resilience will grow in importance right alongside technical AI skills over the next five years.[3] The organisations that get this right early will have something genuinely hard to build later.
You do not need a chief AI officer or a six-figure strategy consultant to begin. Start with a simple audit: where are AI tools already being used in your organisation, even informally? Where could they help most? Which roles would benefit most from even a few hours of AI literacy training?
Bain's research suggests that in the US, up to one in two AI-related jobs could go unfilled by 2027 if companies rely on external hiring alone.[9] The talent market cannot solve this problem at the pace the market needs it solved. The organisations that build from within, consistently and deliberately, are the ones that come out ahead.
Hire curious people. Train the ones you have. Trust that your domain experts are your biggest AI asset. And remember that the companies making the most of AI right now are not the ones with the most sophisticated technology. They are the ones with the most prepared people.
A note from fusecup
At fusecup, we work with businesses building AI-ready teams from the people they already have. If you are trying to figure out where to start with AI adoption, or how to upskill your existing team without waiting for the perfect external hire, we are always happy to have a conversation. No agenda, no pitch. Just a practical discussion about what might work for your specific situation.
The best engineering managers are building towards their own irrelevance, and most get this exactly backwards. They optimise for being indispensable when the actual job is to build a team that runs on shared context, distributed decision-making, and mutual accountability rather than your personal presence in every room.
AI-assisted analytics is spreading fast through data teams. The output looks the same as before. The accountability structure does not.
95% of developers use AI tools weekly. 4.3 million AI repositories exist on GitHub. The data from open source proves that AI-assisted development is not a trend. It is the baseline. The only question left is whether you are being intentional about it.