AI Addiction Check
Below is a plain-language breakdown of how this AI check works, what it measures, and what each question is really asking. Start with the overview of the four areas the check looks at, then read through each question to see what it explores and why it matters. This check is a self-reflection tool only and is not a diagnosis. Your answers stay in your browser and are never stored.
This check looks at how AI chatbots and AI tools fit into your daily life. Researchers who study this describe it as problematic AI use, and the research is still young. It is not an official diagnosis, and heavy use on its own does not mean something is wrong. The check explores whether your AI use feels compulsive or is starting to cause problems.
Control and Time. Does your use run longer than you planned? Do you find it difficult to stop? Do you go back to it after deciding to stop? Loss of control is the central feature of any compulsive pattern.
Emotional Reliance. Do you turn to AI when you feel lonely, anxious, or low? Does the conversation feel easier than talking to people? Do you feel uneasy or irritable without access? These questions look at whether AI has become a coping tool rather than a utility.
Displacement. Does AI use get in the way of your sleep or daily routine? Does it replace time with people, hobbies, or exercise? Do you hide how much you use it? Displacement is where habits start to have a visible cost.
Impairment and Dependence. Has it got in the way of your commitments and responsibilities? Do you rely on it for something you could do yourself? Do you use it even when you would rather not? These mirror the control and consequence questions used in other behavioural addictions.
The check takes about 1 to 3 minutes and is completely anonymous. No signup, no data storage, no judgment.
Losing track of time is one of the earliest signs that a behaviour is becoming hard to steer. AI tools are built to keep a conversation going, so a chat can stretch on without an obvious stopping point. This question asks whether your use outlasts your intention.
Difficulty stopping is the simplest measure of control. Most people can say when they intend to finish. Whether they actually do is a different question, and a more informative one. This item asks how often the stopping point is out of reach.
Going back after deciding to stop takes several forms: reinstalling the app, reopening the tab, or picking the conversation back up. The return is usually driven by the same need that started the habit. This question measures how much staying power your decision actually has.
Using a tool to manage difficult feelings is not automatically a problem, but it becomes one when it is the main way you cope. AI is available at any hour and answers instantly, which makes it a very easy place to land. This question looks at whether emotional relief is part of why you use it.
Chatbots are agreeable, patient, and never tired of you. That can feel easier than a real conversation, which involves effort, risk, and disagreement. When the easier option becomes the preferred one, real relationships can start to lose ground. This question measures that shift.
Withdrawal-like feelings are reported often in surveys of heavy AI users, even though this is not a recognised withdrawal syndrome in the clinical sense. The key detail is whether your mood depends on access. This question explores that link.
Late-night use is one of the most common harms reported by heavy users, partly because the tool never signals that it is time to stop. Sleep loss then affects mood, focus, and self-control, which can feed the pattern. This question looks at the practical cost.
Time is finite, so heavy use of anything comes at the expense of something else. The question is not whether you enjoy AI, but whether other things you value are quietly disappearing. Displacement is often the clearest sign that a habit has become costly.
Secrecy usually appears when someone already senses that the behaviour would be judged, by others or by themselves. It is a useful signal because it is self-generated. This question asks whether shame or defensiveness has entered the picture.
This is the consequence question. Research on students has documented lower academic performance and reduced retention alongside heavy AI use, though the direction of cause is not settled. What matters here is whether you have noticed a cost in your own output or reliability.
Delegation is efficient, but repeated delegation can erode confidence in your own judgement. Several studies describe this as cognitive reliance, where people feel less able to act without the tool. This question looks at whether use has become dependence.
Using something you would rather not use is a clean way to see dependence. Your preference has already been decided, and the behaviour goes the other way. Most people can name at least one thing they do with AI that they would not have chosen to do, and the size of that gap is what this question is asking about.
See a problem on this page? Tell us