Is AI changing how you think?
Most people who reach for ChatGPT ten times a day don't feel "addicted", they feel efficient. This assessment isn't about shame. It's a short, research-grounded check on whether your day-to-day reliance on AI has quietly shifted from a tool you use to a crutch you can't work without.
Takes about 2 minutes. 15 questions, no email required to see your score. If you do leave an email, we'll send your result plus an invitation to the free 5-Day AI Reset: five emails, one small change per day, no quitting AI. If your result suggests it's worth talking to someone, we'll point you to a real resource as well. Your answers are never stored with your address.
What this checks
This assessment borrows from two research instruments published in 2025 and 2026: the AI Addiction Scale (AIAS-21) and the Conversational AI Dependence Scale (CAIDS-20). AIAS-21 is a 21-item scale built around seven dimensions researchers use to describe problematic technology use: compulsive use, craving, tolerance, withdrawal, preoccupation, continuing to use despite harm, and functional impairment. CAIDS-20 is a shorter, 20-item scale focused more narrowly on conversational AI tools like chatbots, covering uncontrollable use, withdrawal, mood modification, and negative impact on daily life.
Neither scale is reproduced here word for word. We adapted the underlying dimensions into 15 plain-language questions that take about two minutes to answer, because most people are not going to sit through a 40-item academic survey on a Tuesday night. The goal was to keep the substance of what researchers actually measure while making it something you would finish. If you want the background on why relying on AI for thinking tasks changes cognitive habits, not just chatting habits, our piece on cognitive offloading covers the research, and what cognitive offloading actually means breaks the term down further.
How the score works
Each of the 15 questions asks how often a pattern shows up in your week, rated from rarely to constantly. Answers are summed into a total score, and the total places you in one of three result bands: Healthy Use, Moderate Reliance, or High Dependence. Healthy Use means AI shows up in your routine but you are not losing sleep, skills, or control over it. Moderate Reliance means AI has started filling gaps you used to fill yourself, in ways worth noticing before they harden into habit. High Dependence means the pattern looks closer to what researchers call compulsive or hard-to-control use: reaching for AI even when you did not plan to, feeling uneasy without it, or noticing it is affecting your work or relationships.
There is no failing score. Most people land somewhere in the middle. The point of the assessment is to name where you are, not to hand out a verdict. If Moderate Reliance sounds familiar, our guide on using AI without losing your skills is a reasonable next stop.
What this is not
This is not a diagnosis. No online questionnaire can tell you whether you have a clinical condition, and this one was not built to try. It is a self-check, the same kind of thing as a stress inventory you might find in a magazine, just grounded in real research instead of guesswork.
Nothing here is stored against your name or your email address. The system logs an anonymous score and result band so we can see aggregate patterns across everyone who takes it, but that record is not linked to who you are. You do not need to give an email to see your result. If you do leave one, it is used once, to send you that result and an invitation to the free course, and nothing more.
Who this is useful for
This is useful if you use AI daily for work or personal tasks and have caught yourself wondering whether it has gone from convenient to compulsive. It is also useful if you manage people and want language for a conversation about AI use on your team, or if you are curious what a research-grounded version of this question looks like next to a viral quiz format.
It is not useful if you are looking for a clinical diagnosis, a substitute for therapy, or a tool built for a child's use. If a teenager in your life is the one you are worried about, that is a different question with different research behind it, and this assessment was not designed for it.
Frequently asked questions
Is this a real psychological test?
No. It's adapted from published research instruments (AIAS-21 and CAIDS-20), but it is a self-check, not a validated clinical diagnostic tool. Treat the result as a prompt for reflection, not a medical finding.
Do you store my answers?
The system logs an anonymous score, result band, and quiz type so we can look at aggregate patterns. It does not log your name, your email, or your individual answers. If you leave an email for your result, that address is used once, for the results email and course invitation.
What happens if I score High Dependence?
You get an honest result and a suggestion, not a diagnosis. The email includes a crisis-line number if you want to talk to someone, along with a link to signs of overreliance at work and the free 5-Day AI Reset course. Nothing about your score gets shared or sorted anywhere.
How is this different from just asking a chatbot if I'm addicted?
A chatbot answering that question has an obvious conflict of interest, and no consistent scale sits behind its answer. This assessment uses the same dimensions researchers use to study AI dependence, applied the same way for everyone who takes it.
Can I retake it later?
Yes. Nothing is tied to your identity, so there is no record linking one attempt to the next. Taking it again in a month is a reasonable way to check whether a change you made actually stuck.
Related reading
- Signs of over reliance on AI at work
- Can't write an email without ChatGPT anymore? Here's what's happening
- AI dependency: how reliance on AI rewires thinking
- AI addiction: what it is, the signs, and what helps