Workers search for ways to hide their AI use from employers

Hidden AI use is rising

Employees are increasingly using artificial intelligence tools at work without approval, and many are actively trying to conceal that use. According to TechRadar, new research from the consultancy Acora shows a marked rise in searches linked to so-called shadow AI, meaning AI used outside corporate oversight.

The study examined Google search behaviour over the 12 months to September 2026. Acora tracked 50 search terms connected with hiding AI activity and found they were entered 161,934 times during that period.

What the search data shows

Acora reports that the tracked terms were searched 48% more often than in 2024, and 250% more than in the first year after ChatGPT launched. The firm says this points to a widening gap between official policy and what staff actually do.

Among the fastest-growing searches are queries about deleting AI chat histories, getting around blocks that employers place on AI tools, and finding free AI services that need no account or sign-up. Some of these searches will come from students rather than workers, TechRadar notes, but the pattern still suggests a growing appetite for tools that operate outside company rules.

Searches for free AI note-taking tools also rose, by 28% on the previous year. TechRadar warns that if such tools are used to record meetings, the notes may end up on remote servers that the organisation has never vetted, creating a data protection risk.

Why policies are falling behind

In principle, staff should know whether AI use is permitted in their role and should have been trained on the relevant rules. TechRadar reports that most organisations do cover this, but the speed of change in AI has left many businesses struggling to keep their policies current.

The result, the research suggests, is that some companies have AI policies that acknowledge the technology without seriously addressing its risks. That leaves room for unapproved use to spread unchecked.

What businesses can do next

Acora's chief executive, David Rabson, argues that companies now face harder questions about data use, trust and whether suitable controls exist. He says concerns about governance and security should not leave organisations trapped between ambition and action.

His proposed remedy is to build proper data foundations, governance and controls so AI projects can proceed with confidence, instead of being paused. For employers, the research is a prompt to make AI rules clear, train staff properly and provide approved tools, so workers have less reason to go around them.