AI Upskilling: Closing the Gap Together

From Job Roles to Future Capability

 
The World Economic Forum’s 2025 Future of Jobs Report found that employers expect 39% of workers’ core skills to change by 2030. According to 2026 research from IDC, over 90% of global enterprises are facing ‘critical AI skills shortages’.

AI is not replacing jobs, it’s reshaping them. And, with this reshaping, a skills gap is emerging.

In the past, we hired for the job. It was enough to ask: “does this person tick the boxes for what we need right now?” Now, we need to hire for potential and capability, acknowledging that the job might not be the same in one year, or five years’ time.

This requires an important shift towards a growth and learning mindset – from ‘filling a gap’ to ensuring our hires are not only able to do their job now, but have the ability to learn and grow in tune with whatever might be coming.

Companies need to understand this job evolution and how that sits in their own market, as well as the potential of AI for their own organisations. To successfully bridge the gap, employers and employees need to be working together on the same path, and not leaving anyone behind.

Taking Everyone With You

I’m a big proponent of people taking responsibility for their own learning. But the fact remains: some groups are consistently left behind in tech transformation. Usually, these are minority groups and women.

This is one of the biggest challenges in AI upskilling, and one that companies need to urgently address. Simply – in order to ensure AI adoption is consistent, long-term and successful – the needle needs to move for everyone. Failure to do so risks job losses and becoming uncompetitive. The talent pool for AI is already too small: if you’re leaving people out, you make it even smaller.

Companies need to look closely at why some of their people are getting left behind.

• Is it because they have no interest in the technology, or because they’re simply not being exposed to it?
• Is there a psychological safety issue, where some people are afraid to lose their job and don’t feel safe to learn?
• For women in particular, whose ‘unseen workload’ is well-documented (BBC, University of Bath, HuffPost) – is it a case of insufficient time, if their colleagues are brushing up on AI outside of work hours?

AI upskilling programmes need to be designed with everyone in mind, because what starts with a little bit less access, or a little bit less time to learn, grows exponentially with time and the speed that the technology is changing.

As a leader, you need to ensure development opportunities are accessible, relevant and within work hours. You also need to make space for your own learning, too, and be aware of change fatigue in yourself as well as your teams.

Bridging the AI Expectation Divide

 
According to research from Tech Radar, 81% of employees believe employers should upskill staff in AI, but 83% of CEOs think employees should train themselves. This is a typical example of the ‘expectation gap’ around AI, and the kind of finger-pointing I see in my work on a daily basis!

The truth is: AI upskilling is everyone’s responsibility.

• Leaders are responsible for explicitly defining and managing expectations and boundaries, outlining what they want to achieve with AI, and building a clear strategy for bringing all employees along on the journey. They need to ask: What do we want, as a company, to achieve with AI? What makes sense from a business perspective, and from an outcome perspective?
• Employees have a responsibility to keep themselves up to date and take ownership of their own learning. This can (and should) include taking issues and concerns to leaders and managers, so AI upskilling can happen in a way that’s relevant and meaningful. Leaders and Boards are great at seeing the bigger picture, but employees have a vital role to play in keeping AI transformation grounded in the day-to-day reality of their organisation. No-one knows it better.

AI is moving fast and no one has all the answers – it’s a learning journey for everyone at the same time. The more employees and employers are aligned, the better the outcome will be.

In some cases, employees can be resistant to change management and fear around AI job-replacement can make this reluctance more acute. Leaders have a pivotal role to play in providing clarity on the end goal of AI adoption, and communicating a framework that answers questions like:

• Why are we implementing AI?
• How do we implement it?
• What do we change?
• What are the challenges?

Traditional change management models move from a stable “current state” to a fixed “future state”. But, because AI capabilities and their impact on work evolve constantly, leaders will find these traditional models insufficient. Change management in AI requires a shift from managing a one-time “go-live” event to managing continuous, unpredictable, and rapid evolution.

The biggest challenge for leaders now is not technology – it’s people, culture and communication. The technology is there, it’s available, it’s been there for a while. But leaders need to decipher what this means from a change management perspective.

Blending Broad and Role-Specific Learning

 
I’m often asked whether generic AI training is sufficient, or if it needs to be specific to job roles and departments. My answer is: there’s a place for both types of training.

Generic training gives people a useful foundation for understanding where AI comes from, what it can do, and how it works. These introductory courses can be a few hours long – I don’t advise investing much more time than that.

Specific AI training tailored to the role and organisation will help you answer the bigger questions fundamental to successful AI deployment. Questions like:

• What do we need from our AI?
• What do you want AI to help you with?
• How are we going to use it?
• How will we apply this knowledge?

This training should incorporate real-case examples relevant to your industry, as well as workflow-based training. Attention spans will wander during even the most interactive training: learning while doing helps embed the information in our brains as well as in our work. The more specific it is, the deeper the learning.

I strongly believe in personalised training content and have, in fact, created a tool to help everyone define their own objectives and support them through their own personalised journeys.

AI Buyer Remorse: When AI Expectations Don’t Match Reality

We’re seeing increasing reports of ‘AI buyer remorse’, with companies back-tracking on massive job layoffs as the expected ROI fails to materialise.

This is, in part, down to poor expectation management and companies failing to identify what they actually want from AI at the outset. AI may be viewed as the new, ‘sexy’ thing (although it isn’t actually new) but rushing headlong towards it without proper planning carries high risks.

Although there will, of course, be areas in any company where AI can optimise staff requirements, mass layoffs are – I think – a missed business growth opportunity to increase revenue and upskill your people. IKEA – who used AI to take on routine customer queries and move thousands of employees into higher-value roles, offers a great example of how AI can be used to remove friction from work, not purpose from people, and unlock significant new revenue at the same time.

Fear, FOMO and AI Anxiety

There’s a lot of fear and ‘push’ around AI. My own social media feeds are full of paid influencers and corporations saying ‘hey, look at this new tool’, and it feels like the hype is constant. The fear of missing out is real, and a growing number of professionals are experiencing AI-related anxiety (Bloomberg).

One misconception that fuels the fear is this:

“technology’s moving so fast, any training will be quickly out of date”.

The first part, about technology evolving quickly, is true. The second part is not. Here’s why:

• Firstly, although the tech is growing rapidly, the new models are not accessible to the majority of us.
• Secondly, companies aren’t embedding the new versions, because they’re still working on fine-tuning existing models.
• Finally, AI training is just one element of the learning required for a job, and it’s a relatively tiny piece. The important thing is understanding how the role might evolve, and to keep learning.

It’s worth bearing in mind that, currently, we only use a small percentage of AI’s capabilities – mostly focussing on the large language models (LLMs) like ChatGPT and Claude. In this way, we’re already ‘behind’, but that doesn’t mean things aren’t evolving!

We all – employers and employees – have a role to play in managing this collective anxiety and fear. In my own learning, I make a conscious effort to read news from the ‘positive’ and ‘negative’ ends of the AI spectrum, as well as the bigger social and environmental picture.

What Next?

When we talk about AI upskilling, it’s important to remember that AI is not new. It has, in fact, been the invisible backbone of our daily lives for decades, evolving from the spam filters of the 1990s to the algorithmic power of Google Search in 2001.

We have consistently upskilled alongside these tools, whether by trusting Netflix’s personalized recommendations in 2007, navigating real-time traffic with Google Maps in 2011, or integrating voice assistants like Siri and Alexa into our homes.

By treating AI as a recent invention, we overlook the fact that we’re already experienced users of mature technologies like automated fraud detection and logistics algorithms. The emergence of GPT-style generative AI in 2022 is simply the next step in this long history, shifting the focus from tools that organize information to those that help create it.

AI is here – and it’s been here for many years – but we’re in a transition period. It will evolve, and it’s not only the large language models that can help us. But let’s not be afraid. Let’s not make the anxiety worse by buying into the ‘we’re all losing our jobs’ hype. The right upskilling, a growth mindset and shared ownership are the keys to ensuring we’re all on the journey together.

To discuss keynotes, workshops or dedicated AI & HR training, get in touch with the Mix Works team.

About the Author