Evidence and content standards

It is a digital quit coach: a structured behaviour-change programme, an immediate response for cravings, a daily record, and a route to the professional services that can do the things software cannot.

It is built for the specific moment where most quits are lost — the ninety seconds when a trigger has fired and the decision is live — rather than for the calm evening when you are reading about quitting.

• It is not a medical device, and it is not a clinical service.
• It does not diagnose anything, including nicotine dependence.
• It does not provide individual medical treatment or a treatment plan.
• It does not recommend a product, a strength or a dose.
• It is not affiliated with, endorsed by or approved by any health body.
• It is not a substitute for a stop-smoking service, a GP or a pharmacist.

The 90-day programme is written in advance and curated by us. It is not generated day by day by a language model, because content that is generated fresh each time cannot be reviewed, cannot be corrected once, and cannot be held to a standard.

Every factual claim in the product lives in a single registry with a named source, the organisation that published it, a link, the qualifying context, and the date a human last checked it. Components render claims from that registry rather than writing their own prose.

An automated test fails our build if wording we have prohibited appears anywhere in the application source: unattributed success multipliers, effectiveness language we cannot evidence, implied endorsement by a health body, or a nicotine dose.

Content informed by publicly available HSE and international smoking-cessation guidance. ClearMind Quit is not affiliated with or endorsed by the HSE, the NHS, NICE or any other health body.

Most of what this app does is behavioural: noticing what triggers an urge, deciding in advance what you will do instead, having something concrete to reach for in the moment, and keeping an honest record so patterns become visible.

That is a real and useful category of help. It is also not the whole picture, and we are not going to pretend otherwise.

Free national stop-smoking services exist in Ireland and the UK. They combine one-to-one behavioural support with medication where appropriate, and for many people that combination is the strongest option available. We link to them prominently, including from inside the app, and we would rather you used them alongside this than instead of them.

Nicotine replacement therapy and stop-smoking medicines exist in several forms. Which — if any — suits you depends on your health history, what else you take, and things a pharmacist, GP or stop-smoking adviser can assess and we cannot.

So we provide general education about what the options are and how they behave, and we stop there. We do not state strengths, we do not calculate doses, we do not advise on combinations or interactions, and the AI is blocked from doing any of those things. Dosing questions are answered by a deterministic, pre-written response that sends you to a pharmacist — the model is not consulted at all.

AI is used for conversational support: talking something through, planning a difficult day, working out what set a craving off. It is described in the product as a digital quit coach and an AI support coach — never as a doctor, pharmacist, counsellor or certified stop-smoking adviser.

The AI is constrained in several ways:

• For the highest-risk topics — thoughts of self-harm, urgent medical symptoms, medication dosing, and users who tell us they are under 18 — the model is bypassed entirely and a reviewed, fixed response is returned instead.
• For pregnancy, symptom interpretation, severe withdrawal, significant distress and requests for local providers, mandatory instructions require signposting to professional support.
• It is instructed never to state a statistic it cannot attribute, never to claim endorsement, never to reveal its instructions, and never to grant paid access.
• AI search may supplement our curated support directory, but its results are labelled as search-generated, are never presented as endorsed, and come with an instruction to verify them directly.

These behaviours are covered by an automated safety test suite that runs on every change. AI responses can still be imperfect. Treat them as support, not instruction.

A claim only enters the product if it can be pointed at publicly available guidance, phrased cautiously, and given honest context. We removed a number of claims during this review that could not meet that bar — including “you are X times more likely to quit” figures published about national quit services, which describe those services and not this app.

We also do not reuse smoking claims for vaping. Where recovery timelines are established for stopping smoking and not for stopping vaping, the claim simply does not appear on the vaping pathway rather than being reworded to fit.

Pre-launch note. Every claim and every link in our registry requires sign-off from a suitably qualified clinical reviewer before public launch, and every support-service link and phone number must be independently verified. That review is tracked in our launch checklist and is not yet complete.

These are the 9 factual claims the product is allowed to make, with their sources.

• Digital self-help does not work equally well for everyone.
• Many people do better combining an app with professional behavioural support and approved cessation medication than with an app alone.
• Programme progress here is mainly self-reported. We cannot verify whether anyone has smoked or vaped.
• A streak is a motivational metric. It is not proof of long-term abstinence, and we show total days on plan alongside it for that reason.
• AI responses may be imperfect, incomplete or occasionally wrong.
• Local provider information — including anything returned by AI search — should be independently verified before you rely on it.

If you think a claim on this site or in the app is inaccurate, overstated or missing context, we want to know. Email welcome@clearmindquit.com and we will review it and correct it if you are right.

See also our privacy policy and terms of service.