Fresh from Vilnius: see how we pitched AI Adoption Score to an audience of 800+ at TestCon Europe on October 23Fresh from TestCon, Oct 23: our pitch to 800+ people

Safe AI Use Quiz

Being sure isn't knowing. Here's how we check whether people know what's safe to do with AI tools at work: a self-rating and a quiz in 7 areas, with the level set by the quiz.

  • 7areas, from tools and accounts to AI scams, each with a self-rating and three quiz questions
  • 21quiz questions with one right answer; the level comes from them
  • 4questions on rules and habits, not scored
  • 15 minper person, on a phone or a computer
Start Safe AI Use Quiz

Free to start, live in two clicks.

Step 1 of 5People rate themselves in seven areas first, so the quiz can't change their rating.

What is the Safe AI Use Quiz report?

A check of whether people know what's safe to do with AI tools at work, whatever tools they use. Everyone rates themselves in 7 areas, from tools and accounts to AI scams, then answers three short questions in each, in about 15 minutes.

The level, from Beginner to Expert, comes from the right answers, not from how sure people are. The report shows the gaps by area and by team, who is sure but wrong, how many people use their own AI accounts for work, and whether they know your AI rules.

The short version

Everyone at work, not only people who use AI every day, rates themselves in 7 areas and answers three short quiz questions in each. Every question has one right answer, so the level comes from what people know, not from how sure they are. Four questions on your rules and their habits come first and aren't scored. The whole thing takes about 15 minutes.

Each person gets a level and a plain tip for every area where they missed a question. The report shows gaps by area and team, where people are sure but wrong, and what they say about your rules and their habits. It is not a certification.

Seven areas

Built from data-protection guidance, security frameworks and real incidents with AI tools from 2023 to 2026. The questions test lasting principles that hold for any AI tool, not one vendor’s settings. Each area opens with the same self-rating question, then its three quiz questions.

Tools and accounts

Approved tools and work accounts; personal plans, free apps and browser extensions

Quiz questions
3, one each: basic, practical and advanced.
Self-rating
"How would you rate yourself in choosing AI tools and accounts for work?", from Beginner to Expert.

The area's self-rating

How would you rate yourself in choosing AI tools and accounts for work?

Approved tools and work accounts; personal plans, free apps and browser extensions

Pick a level. People rate themselves before the quiz, so the questions can't change the rating.

Before the areas, About you asks for a team (optional, from a list set per round), "How often do you use AI tools for work?" (Rarely or never, Monthly, Weekly, Most working days) and "Which accounts do you use for AI tools at work?" (One my company gives me, My own free account, My own paid account, I don’t use AI tools for work; several allowed), so the report can set use and accounts against the score.

A quiz, not a self-rating

Self-ratings measure confidence. People who are sure and wrong are the ones an opinion survey misses, and the ones most likely to paste a customer list into the wrong tool. So the level comes from a quiz: three questions per area, one each at a basic, practical and advanced level, each with four answers, one of them right, plus "I’m not sure". A right answer is 1 point; a wrong one, or "I’m not sure", is 0.

The real questions aren't published, so the quiz stays fair from one round to the next. Here is one of ours in the same shape, with its answer:

Where data goesPractical

Your AI assistant has memory switched on. A month ago you used it to work on a client's prices. What can happen now?

Pick an answer to see whether it's the right one, and why.

  • Work situations, not definitions. Each question describes something that happens at work: a browser extension that asks for everything, a share link, a video call that asks for an urgent payment.
  • Being too careful is a mistake too. In some questions the bolder answer is right, such as using the assistant your company approved for the internal work it's meant for, so the quiz also finds people who avoid the tools they may use.
  • The right answers stay on the server. They never reach the survey, and people never see them, not even in their own result. Whoever can open the report sees them, ticked, on the Questions page.
  • No tool, plan or price is named, so the questions hold whatever tools your company uses. When a right answer has to change, the question gets a new identity, so rounds compare only questions that stayed the same.

Rules and habits

Four questions come before the quiz and aren't scored. They come first, so the quiz can't sway the answers:

  • Does your company have rules for using AI tools at work?

    Yes, written downYes, but not written downNo rulesAI tools are bannedI’m not sure

  • Have you had training on the risks of AI tools?

    Yes, in the last 12 monthsYes, longer agoNoI’m not sure

  • In the last 3 months, how often did you put company or customer information into an AI tool your company hasn’t approved?

    NeverOnce or twiceA few times a monthWeekly or moreI don’t know which tools are approved

  • If you shared something you shouldn’t have with an AI tool, would you know whom to tell?

    YesI’d find outNo

Rules and habits: not part of the score.

The report shows how many people gave each answer and their average score: whether people who had training score higher, how many aren't sure your company has AI rules, and how many use tools you haven't approved, or don't know which ones are. Whoever launches the survey decides whether it runs anonymously, as with every survey.

From right answers to a level

The score is right answers out of 21, and the level comes from it. The same cut-offs apply to everyone, whatever they rated themselves.

  1. BBeginner0–6 rightYou are just starting to learn what’s safe with AI tools.
  2. EExplorer7–10 rightYou know some rules, but common traps still catch you.
  3. PPractitioner11–14 rightYou handle most everyday situations safely.
  4. AAdvanced15–17 rightYou spot the less obvious risks too.
  5. EExpert18–21 rightYou know what’s safe in almost every situation and can help others.

Move the score to see the level it gives:

P

Practitioner

You handle most everyday situations safely.

An area is scored on its own too, as right answers out of 3.

Confidence against results

The self-rating isn't scored. It's set against the quiz, area by area, to show where confidence and results part ways:

Blind spot

Someone who rated themselves Advanced or Expert in an area, but got at most 1 of 3 right: sure, but wrong. They see "You rated yourself higher than you scored" for that area, and the report counts them in the area's blind spots. That's where training should start.

Hidden strength

Someone who rated themselves Beginner or Explorer in an area, but got 3 of 3 right. They see "You know more than you think" for that area. A Practitioner rating is neither, whatever the quiz says.

In the report's Confidence and quiz section, each area shows how people rated themselves next to the share of answers they got right.

What people see

Whoever launches the survey decides with one switch, Show people their own results, on by default. Off, people see a thank-you message only.

  • Their own result, right after submitting: the level with its one-line summary, each area as right answers out of 3, and where their self-rating and the quiz disagree. With it a private link they can come back to.
  • Tips for you: one plain rule for each area where they missed a question, so the quiz teaches as well as measures. A tip never says which answer was right.
  • What they never see: the right answers, other people's results, or the company's numbers.

What the report shows

The report fills in from the first response and has three pages: Overview, Questions and People.

The Overview page of a sample report, with fictional data.
  • Top summary: the average score out of 21 with the share of answers right, how many people are at Practitioner or above, the biggest gap (the area with the lowest share of right answers), and how many people took it.
  • Overview: how many people reached each level; the share of right answers in each area; a heat map of teams by area; Confidence and quiz; Use and skill, the average score by how often people use AI tools and by the accounts they use; and Rules and habits, each answer with its share of people and their average score.
  • Questions: every question with the share that picked each answer, the right one ticked, and the most common mix-up: the wrong beliefs to teach against first.
  • People: everyone who answered, sorted by name, never ranked, each with their score, areas and blind spots. Open a person for the result they saw.

The Trends tab fills in from the second round: the average score and the share at Practitioner or above, each round's key numbers and biggest gap, the levels, the three lowest areas and the average score by how often people use AI tools, round by round, for the same version of the questions. Run it again after training to see whether it worked. The report compares only with the earlier reports you pick for it, such as the same team's last rounds: how Trends work.

The Trends tab of a sample report, compared with its earlier rounds, with fictional data.

Who sees what

  • It can run anonymously. The launch window starts with a shared password and no names. You can ask for a name or an email instead, or have people sign in with Google or an emailed link, and then the report names people, with the People page and Responses open to whoever can open the report.
  • Nobody is ranked. People are sorted by name. The public link shows the Top summary and Overview only, and the AI connection returns aggregates, never answers or names.
  • For learning, not for decisions about people. As the survey's own guidance says: "Results are for learning and planning, not for hiring or firing decisions on their own."

Limits

  • It is not a certification. It isn't supervised, and three questions per area is a quick check, not a full examination: it tells you where a team stands and where to start, not whether one person is safe.
  • Knowing isn't doing. Someone can know the rule and still break it. The habits question shows what people say they do; only tools that watch real traffic show what they do.
  • Your own rules come first. Which tools your company approved and whom to tell differ from company to company, so the quiz tests principles, and the rules and habits questions ask about yours.
  • Small groups swing. Results show from the first response, with no minimum group size, so a small team's numbers can point to one person.

Don't renew your
AI licenses blind.

Before the renewal, see which teams get value from AI,
which don't, and why.