
Before trusting a new AI app, check its privacy policy, data controls, permissions, pricing, company credibility, user reviews, safety boundaries, and whether its claims sound realistic. Do not rely only on marketing language.
Start with privacy. What data does it collect? Can you delete it? Is memory optional? Is your data used for training? Then check permissions: does the app ask for contacts, files, location, or calendar access before it clearly needs them?
Also test the app with a low-risk task first. See whether it gives useful answers, admits uncertainty, and avoids overpromising. Apps that make extreme claims about health, money, relationships, or guaranteed results deserve caution.
Run one small test before trusting: tell the app a harmless detail, then find it, edit it, and delete it through settings. An app that passes this loop cleanly has earned the next level of information.
Order your checks by stakes: first whether the company states it does not sell personal data, then whether memory is user-controllable, then whether deletion is real. An app must pass all three, in that order.
Your checklist should end with a personal rule, not just app facts. Decide before the first session what you will never enter, such as identification numbers, credentials, or other people's private information.