Why AI Is the UK's Biggest Tech Skills Gap in 2026, and How to Close It

The key findings at a glance

• AI is the top technology skills gap. 35% ranked it the single biggest gap in their organisation and 70% placed it in their top three, well ahead of cyber security in second place.
• Shadow AI is widespread. 42% use AI tools at work that their organisation has not provided or approved: 20% regularly and 22% occasionally.
• Retraining is falling short. Only 11% rate their organisation's AI retraining as very effective, and nearly 30% say no retraining has happened at all.
• Adoption is rated as middling. Respondents scored their organisation's AI adoption at 2.84 out of 5, and their confidence that AI is being used ethically and responsibly at 3.28 out of 5.
• Trust and security are the biggest barriers. Accuracy and trust concerns (59%), security and data protection (59%) and a lack of skills (49%) top the list.
• Upskilling is the preferred fix. 75% placed upskilling existing staff in their top three ways to close skills gaps.
• Support for recognised ethics training is strong. 52% strongly agreed that anyone using AI at work should hold recognised ethical training or certification.

The wider picture points the same way. Government research cited in the report found that only 21% of UK employees feel confident using AI at work, at a time when the government's AI Skills Boost programme aims to give 10 million workers foundational AI skills by 2030.

Why the gap is so hard to close

People are teaching themselves

A recurring theme in the report's written responses is that staff have been left to work AI out as they go. Self-directed learning has real value, but it is uneven. Two people in the same team can end up with very different levels of understanding, and nobody is quite sure what good practice looks like.

Training covers the tool and skips the judgement

Where training does exist, respondents describe it as shallow. One summed up their organisation's AI training as simply learning how to use Copilot. Knowing which buttons to press is a start. The harder skill, and the one respondents say is missing, is knowing when to trust an AI output, how to check it, and when AI is the wrong tool for the job.

Adoption is moving faster than governance

Shadow AI is the clearest sign of this. When more than two in five tech professionals use tools their employer has not approved, it shows a real appetite for AI. It also means sensitive information can end up in systems nobody has assessed. The report recognises both sides: the demand is genuine, so the task for leaders is to govern it well.

Leaders are short of skills too

The second most cited skill gap was the ability to automate processes and redesign workflows around AI, named more often than prompt engineering. Closely related is the ability to judge where AI genuinely adds value, and more than a quarter of respondents named a lack of clear use cases as a main barrier to adoption. These are leadership decisions, and the report suggests many senior teams need more support to make them with confidence.

The AI skills organisations need most

The report groups the skills in short supply into four areas:

1. Building AI systems, with agentic AI development the single most cited skill.
2. Applying AI to everyday work, putting AI to use in day-to-day roles.
3. Governing AI, making sure AI is used safely and responsibly.
4. AI literacy and judgement, the confidence to use AI well and the critical thinking to question what it produces.

Most organisations need a small number of specialists in the first area and a far larger number of people with skills in the other three. That second group is where structured, accessible training makes the fastest difference.

How to close the AI skills gap: five practical steps

1. Build a shared baseline of AI literacy

Start with a common understanding across the team: what AI is, how machine learning and generative AI work, where the benefits and risks lie, and how to check outputs. A recognised entry-level course such as the BCS Essentials Certificate in Artificial Intelligence or the BCS Foundation Award in Generative AI gives everyone the same foundations and the same vocabulary.

2. Train for judgement as well as tools

Tool training has its place, but it dates quickly as products change. Principles last. Look for training that teaches people to question, verify and apply AI, which is exactly the gap the BCS report identifies.

3. Make ethics part of the plan from day one

The strongest single call in the BCS research was for recognised ethics training for anyone using AI at work. Ethics in this sense is practical: fairness, transparency, accountability and clarity over who is responsible when an AI-assisted decision goes wrong. The BCS Foundation Award in Understanding the Role of Ethics in the Responsible Use of AI covers this ground in depth.

4. Bring shadow AI into the open

Find out which tools people are already using and why. Approve the ones that meet your security and data standards, provide alternatives where they fall short, and give staff clear guidance on what can and cannot be shared with an AI tool. The aim is open, supported use that leaders can see and manage.

5. Give leaders the skills to spot real value

Managers and senior leaders need enough understanding to identify strong use cases, redesign workflows and make sound procurement decisions. The BCS Foundation Award in How AI Can Support Your Organisation helps people understand the role AI can play in shaping their organisation's future. For those who want a broader, deeper grounding, the BCS Foundation Certificate in Artificial Intelligence builds on the essentials with more practical insight into machine learning and how AI is applied in organisations.

A note for the public sector

The report highlights particular pressure in the public sector, where respondents linked unfilled gaps to tight budgets, recruitment freezes and pay that struggles to compete. When hiring specialists is difficult, upskilling the people already in post becomes the most realistic route, and flexible online training that fits around the working week makes that far easier to deliver.

What to look for in an AI course

• A recognised, independently assessed certificate. BCS professional certificates are set and examined by the professional body for IT, so they carry weight with employers in the UK and internationally.
• Content that covers ethics, risk and judgement alongside how AI works.
• Flexibility to study around work, at your own pace.
• Tutor support for questions as they come up.
• Clear pricing, with the exam included.

How Duco Digital Training can help

Duco Digital Training is a BCS-accredited training provider specialising in AI, with learners across the UK, Europe, the Middle East and South East Asia. Our AI courses lead to BCS professional certificates and focus on practical application at work. Our lead AI trainer is Darren Winter, our CEO, who is completing a PhD in AI ethics.

Every course includes:

• the BCS exam as standard
• 12 months' online access, so you can study at your own pace
• tutor support throughout, including on WhatsApp
• Pass Assist exam preparation support
• a free year of BCS Associate Membership when you pass, for new BCS members

Whether you're building AI skills across a team or taking the next step in your own career, get in touch on WhatsApp or through our website and we'll help you find the right course.

Frequently asked questions

What is the biggest tech skills gap in the UK in 2026? According to the BCS AI Skills and Adoption Report 2026, AI is the biggest technology skills gap identified by tech professionals, for the second year running. 35% ranked it first and 70% placed it in their top three.

What is shadow AI? Shadow AI is the use of AI tools at work that an organisation has not provided or approved. In the BCS research, 42% of tech professionals said they do this, either regularly or occasionally.

Which AI skills are most in demand? The report identifies four areas: building AI systems, applying AI to everyday work, governing AI, and the literacy and judgement to use AI well. The single most cited skill was agentic AI development, followed by process automation and workflow design.

How effective is workplace AI training? The report shows plenty of room for improvement. Only 11% of respondents rated their organisation's AI retraining as very effective, and nearly 30% said no retraining had taken place at all.

Should everyone who uses AI at work have ethics training? The BCS research shows strong support for it. Asked whether anyone using AI at work should hold recognised ethical training or certification, respondents averaged 4.06 out of 5, and BCS recommends a recognised standard for responsible AI use.

What is the best AI certificate for a non-technical professional? For most non-technical professionals, the BCS Essentials Certificate in Artificial Intelligence is a natural starting point. It is an entry-level introduction to the principles of AI, its benefits and risks, and ethics. Those who want more depth can choose the BCS Foundation Certificate in Artificial Intelligence.

Source: BCS, The Chartered Institute for IT, AI Skills and Adoption Report 2026 (September 2026).