Most organizations discover that employees use far more AI tools than IT has approved. This survey template gives you 30 well-crafted questions, a ready-to-send amnesty message, and an analysis guide with a simple exposure score, so you can build your first honest baseline of AI use and the data flowing into it.
Blocking and monitoring only show you traffic. A well-run survey shows you motive: which tasks drive AI use, why employees bypass approved tools, and where your policy has not landed.
Employees answer truthfully when leadership commits, in writing, that nothing disclosed will be used against them. The template bakes this no-blame framing into the introduction message and the survey design itself.
Personal accounts, browser extensions, meeting bots, and AI features inside approved software: the usage that network logs miss, plus the unmet needs and blockers that no technical control can reveal.
Surveys capture what people remember and admit on one day. Usage they consider trivial, tools they forgot, and everything adopted after the survey closes stay invisible until the next round.
Thirty questions in eight sections, each marked as single choice, multi-select, Likert scale, or free text. Role questions stay deliberately coarse so small teams cannot be identified.
Multi-select lists of common assistants, writing and design tools, coding tools, and meeting notetakers, each with an open field for anything not listed.
How often AI is used, for which tasks, and how much time it saves, so you can size adoption rather than guess at it.
Classification-based options from public information through to customer data and credentials, plus how carefully content is checked before it goes in.
Personal versus company accounts, unsanctioned sign-ups with work email, and standing integrations connected to company systems.
AI browser extensions on work browsers and AI notetakers in internal and external meetings, two of the most overlooked exposure paths.
Broad function groups and manager status only, deliberately coarse to protect anonymity while still enabling segmentation.
What stops employees from using approved tools, and which task they most want a better tool for: your procurement shortlist.
Whether employees know the AI policy exists, how to get a tool approved, and how comfortable they feel being open about their usage.
The template includes a full analysis guide: how to segment results without breaking anonymity, which answer combinations to treat as red flags, and how to compute a simple shadow AI exposure score out of 12.
Each response is scored across data sensitivity, account type, usage intensity, standing access, and awareness gaps, then averaged into an organization-level score.
Maintain guidance and re-survey in 12 months.
Prioritize the top red-flag segments this quarter.
Launch a formal shadow AI program within 30 days.
Treat as an active data-exposure incident and brief executives now.
A worked example is included: a filled summary table for a hypothetical 420-person organization, showing how findings, score inputs, and priority actions fit together.
A survey is your baseline; it goes stale the day it closes. Aona discovers every AI tool in use across your organization continuously, shows what data flows into each one, and keeps your shadow AI picture current between surveys.