It & telecoms workers most likely to use shadow AI

TrustedTech has published research showing that employees in IT and Telecoms are the most likely to use unapproved AI tools at work. The survey covered 2,001 employees in the UK and US.

The data points to a gap between awareness of AI risks and workplace behaviour. In IT and Telecoms, 73% of respondents said they use shadow AI tools, while 84% said they recognise the security or data privacy risks tied to them.

Finance ranked next, with 59% of employees reporting unapproved AI use and 80% saying they understand the risks. HR followed at 58%, with Architecture, Engineering and Building at 54%, and Manufacturing and Utilities at 52%.

Lower use in some sectors did not remove the issue. Healthcare recorded 41% unapproved AI use despite 76% risk awareness, while Education was lowest at 38%, with the same 76% level of awareness.

The findings suggest that knowing the risks does not necessarily stop employees from turning to unauthorised tools. Across all sectors measured, awareness levels were consistently higher than the share of workers who said they avoided shadow AI.

Rule breakers

The survey also examined whether workers would continue to use AI tools if their employer banned them and disciplinary action was possible. IT and Telecoms again ranked highest, with 48% saying they would continue, followed by Finance at 41%, Architecture, Engineering and Building at 36%, and HR at 34%.

At the other end of the scale, Travel and Transport recorded 20%, while Education and Retail, Catering and Leisure both stood at 21%. Those figures suggest a lower willingness in those sectors to ignore formal restrictions.

The pattern was especially notable in sectors that often write or enforce technology rules for the rest of the business. Workers in IT functions can also have broader access to systems, organisational data and administrative privileges, which may increase the consequences of using external or unapproved tools.

Guidance gap

The research linked higher shadow AI use to weak internal guidance and limited training. In IT and Telecoms, 57% said their organisation lacks clear guidance on using AI at work, while 58% said they do not receive enough training to use those tools safely and effectively.

Similar issues appeared elsewhere. Half of respondents in Architecture, Engineering and Building said guidance was lacking, and 49% in HR said the same. On training, Manufacturing and Utilities and Finance both recorded 48% saying provision was insufficient.

The study also found a link between use and perceived time savings. More than half of IT and Telecoms workers, 52%, said AI saves them five hours or more each week. That compared with 20% in Healthcare and 19% in Education.

Those differences suggest that sectors with heavier day-to-day digital workloads may feel a stronger pull towards external AI tools when approved options are absent, restricted or less useful for routine tasks. The data also shows that sectors with lower reported use still face meaningful exposure where sensitive information is involved.

Workplace culture

Beyond policy and training, the survey highlighted tensions over how AI use is viewed by colleagues. In both IT and Telecoms and Sales, Media and Marketing, 28% said they judge co-workers negatively for relying heavily on AI in their workload. In Retail, Catering and Leisure, that figure was 12%.

Self-censorship was also most evident in IT and Telecoms. Some 31% said they reduce their own use of AI because of how they may be perceived by colleagues or managers, ahead of HR and Manufacturing and Utilities, both at 26%.

The figures add to a broader debate over how companies govern workplace AI as use spreads faster than formal policy. They also raise questions for heavily regulated sectors, where employees may handle financial records, customer data, health information or internal systems while using tools that sit outside approved controls.

In a statement accompanying the research, Julian Hamood, Founder and Chief Visionary Officer at TrustedTech, said: "There is an assumption that Shadow AI is a problem of ignorance - that people use unapproved tools because they do not understand the risk. Our sector data shows the opposite. The industries that understand AI best are the ones bypassing the rules most often, and the most willing to do so knowing it could cost them their job. That should reframe how organisations respond. This is not a training problem to be solved with another awareness module - it is a provision and policy problem. Where approved tools are slower, narrower or more restricted than the alternatives, employees will route around them, and the more capable the workforce, the faster they will do it. Sectors handling regulated or highly sensitive data, particularly Finance and Healthcare, cannot afford to treat that as an acceptable trade-off."