Blog · AI Upskilling
Shadow AI at work: why are UK staff using AI without telling their employer?
· By Day Seven
Short answer: New research from Imperial Lifelong Learning finds that 29% of Gen Z workers in high-growth UK sectors use AI without their employer knowing, while separate Howden research shows half of workers worry AI is arriving faster than they can adapt. Shadow AI is usually a sign that people want to use AI but lack clear rules, approved tools and training, so the fix is to bring it into the open rather than ban it.
What the new research says
On 11 November 2026, Imperial College London launched Imperial Lifelong Learning with a report on how AI is used and viewed at work. Censuswide surveyed 2,000 UK employees across five high-growth sectors between 4 and 15 September 2026. Imperial runs courses for working adults, so it has an interest in the topic, but the findings are worth reading.
According to Imperial's press release:
- 29% of Gen Z workers are using AI without their employer's knowledge, compared with 18% of Millennials, 11% of Gen X and 6% of Baby Boomers.
- 35% of Gen Z respondents say using AI at work feels dishonest, and 26% worry colleagues will view them negatively for using it.
- 75% of employers recognise a significant skills gap in their industry, yet 97% of workers are confident they have the skills their role will need in the next 12 months.
The report page adds that 85% of employees want further AI-related training in the next two years. Fathom, which reported on the full report, says 17% of all workers surveyed use AI without their employer knowing, and that 20% avoid AI because they do not trust the results.
The same day, Workplace Journal and DIGIT reported research from Howden Employee Benefits. It found that 50% of UK workers worry their employer will introduce AI faster than staff can understand or adapt to it, 53% are concerned about finding time to learn new skills alongside their workload, and 50% say their employer gives them no protected time each month for learning and development.
Why shadow AI happens
Put the two studies together and a pattern appears. Many people are already using AI, some of them quietly. Others hold back because it feels like cheating or they do not trust it. And many say they have no time set aside to learn.
None of that points to staff who do not care. It points to a gap between what people are doing and what their organisation has made clear. When there is no approved tool, no plain policy and no training, people either work around the rules or avoid AI entirely. Both are a problem. The first creates risk, because nobody knows what data is going where. The second leaves value on the table.
The confidence finding matters too. If almost every worker thinks their own skills are fine while most employers see a gap, optional training will struggle to fill seats. Training has to be tied to real work so people can see what they are missing.
Practical takeaways by role
L&D leaders: Build training around the tasks people already use AI for, such as drafting, summarising and analysing data, rather than generic tool tours. Include how to check outputs, since some staff avoid AI because they do not trust the results. Make sessions part of protected learning time, not something squeezed in after hours. Our AI upskilling for staff is designed around this kind of practical, role-based work.
HR leaders: The finding that many Gen Z workers feel using AI is dishonest is a culture issue as much as a skills one. Tell people clearly when AI use is expected and welcome, and when it is not. Check that your employee AI policy is short enough to be read and covers which tools are approved. Our AI training for HR teams can help you lead this.
IT leaders: Shadow AI usually means people could not find an approved tool that does the job. Find out which tools are in use, without blame, and make sure there is a sanctioned option that is easy to reach. Set simple rules on what data must never go into an unapproved tool, and explain the reason behind them.
Transformation leaders: Howden's research suggests staff are watching how the gains from AI are shared. Be open about what you are trying to achieve, involve teams early and measure the effect on real work. Our guide to measuring AI training ROI shows how to set a baseline before you start.
A short checklist
- Ask first. Run a quick, anonymous survey on which AI tools staff use and for what.
- Approve something. Make sure there is at least one approved AI tool people can use for everyday work.
- Publish clear rules. Keep the AI policy to a page and state what data stays out.
- Train on real tasks. Use people's own work as the training material.
- Protect the time. Set aside learning time each month so training is not an extra burden.
What this means for your team
If some of your people are using AI quietly, that is useful information, not a reason to clamp down. It shows where demand is and which tasks AI already helps with. The aim is to make good AI use visible, safe and normal, so nobody feels they have to hide it and nobody is left behind because they were never shown how.
You can read more in our earlier posts on the AI confidence gap between leaders and staff and PwC's findings on daily AI use and training access, or find out more about our corporate AI training, AI upskilling for staff and AI training in Leeds.
Talk to us
If you would like help bringing AI use into the open and giving your teams the skills to use it well, book a free 60-minute discovery call with Qasim.
Frequently asked questions
What is shadow AI?
Shadow AI is when staff use AI tools for work without their employer knowing. Imperial Lifelong Learning's new report defines it as using AI without your employer's knowledge, and finds it is most common among Gen Z workers.
How many UK workers use AI without telling their employer?
Imperial Lifelong Learning's survey of 2,000 UK employees in five high-growth sectors found that 29% of Gen Z workers do so, compared with 18% of Millennials, 11% of Gen X and 6% of Baby Boomers. Fathom, which reported on the full report, says the figure is 17% across all workers surveyed.
Should we ban AI tools to stop shadow AI?
A ban on its own tends to push use further out of sight. A better approach is to provide approved tools, publish a short and clear AI policy, and train people in how to use AI safely for their real tasks, so there is no reason to hide it.
What do employees want from AI training?
Imperial's report says 85% of employees want further AI-related training in the next two years. Howden Employee Benefits research, reported by Workplace Journal, found that 53% worry about finding time to learn new skills and 50% say their employer gives them no protected time each month for learning.
Where should an employer start with shadow AI?
Start by asking staff, without blame, which AI tools they already use and for what. Then confirm which tools are approved, set simple rules on what data can go into them, and run practical training built around the tasks people are already using AI for.
Related posts
AI skills divide: why are senior leaders getting more from AI than junior staff?
A new PA Consulting survey of 2,000 UK employees finds senior leaders are getting far more value from AI than junior staff, and that faster work does not always mean better work. Here is what that means for L&D, HR, IT and transformation leaders.
Daily AI use is rising at work, so is employee AI training keeping up?
PwC's Global Workforce Hopes and Fears Survey 2026 finds that daily generative AI use has risen to 22% of workers, while fewer say they have the learning they need. Here is what that gap means for L&D, HR, IT and transformation leaders planning AI training.
AI upskilling: what can employers learn from the government's Barnsley training projects?
On 3 November 2026 the government awarded over £800,000 from the AI Upskilling Challenge Fund to four practical AI training projects in Barnsley. Here is what L&D, HR, IT and transformation leaders can take from the way those projects are designed.