Character chat invites candour: people write personally to characters in a way they would not post publicly, which makes privacy and account safety more important here than in most apps. This guide covers the whole surface for Janitor AI for Android phones: account security, the special care API keys demand, what to remember about chat privacy, and the community tools — blocking, reporting, and content controls — that keep the platform comfortable.
Nothing here is complicated, and all of it is cheaper than the incident it prevents. Read it once, apply it in twenty minutes, and then forget about it the way you forget about a locked door.
One principle organises everything below: fiction should stay fiction. The stories are invented; the account, the credentials, and the person are real, and they deserve real-world protection.
Account Basics, Done Properly
Use a unique password for this account, stored in a password manager, and keep the account's email current and itself secured. A chat account holds your characters, your history, and your configurations; it deserves the same tier of care as the email it is attached to.
Where additional verification is offered, enable it. Account recovery is decided on the day you set it up, not on the day you need it, and the difference between those days is the whole story of most lockouts.
API Keys Are Money and Identity
If you use your own API key, treat it as both a password and a payment instrument: never share it, never paste it into unfamiliar services, and revoke it at the provider the moment you suspect exposure. A leaked key can spend your money in someone else's conversations.
Beware services offering impossibly free access in exchange for your key or your traffic. The settings guide in this library explains the legitimate structure; anything that does not match that structure deserves your distance rather than your credentials.
Chat Privacy, Understood Clearly
Know, in general terms, what the platform and any connected model provider can process: your messages travel to a model to be answered, and providers' policies govern retention. Read the current privacy documents once, with your actual usage in mind, rather than assuming either total exposure or total secrecy.
Write accordingly. Keep real names, addresses, workplaces, and identifying details of real people out of chats, yours included. Fiction loses nothing by the omission, and privacy gains everything. The persona guide in this library makes the same point from the craft side: invented details serve the story better anyway. See the Janitor AI persona settings guide for personas built entirely from invention.
Content Controls and Comfort Settings
Platforms like this one serve adult audiences with content controls for a reason. Set the filters and preferences to your own comfort deliberately rather than inheriting defaults, and revisit them when your usage changes. Your feed of characters is a garden; the controls are its fence.
If you share a device, remember that comfort settings protect the next pair of hands as well. A signed-in app on a shared phone is a household decision, and households deserve a moment of joint configuration.
Blocking, Reporting, and Community Hygiene
Use block and report without hesitation when a user or a character crosses into harassment, spam, or content that violates the platform's rules. Reporting is maintenance, not drama: it is how a large creative commons stays usable for the people behaving well inside it.
Do not engage with provocation to defend your work. Screenshot if evidence matters, report, block, and return to your stories. Attention is the currency trolls spend; the block button is how you stop paying. The discovery guide in this library shows where community content comes from — finding characters in Janitor AI — and hygiene is what keeps that commons worth searching.
Your Characters and Your Rights
Original characters you create are your creative work. Publish them with the awareness that public characters can be imitated in spirit, and keep your master notes — the full definitions and private lore — in your own files, publishing only what the chat needs.
If you believe your work has been copied improperly, use the platform's reporting channels with specifics: links, dates, the passages or traits at issue. Specific reports get acted on; general grievances mostly get archived.
A Twenty-Minute Safety Setup
Assemble the whole page into one sitting: unique password stored, recovery email verified, extra verification on, keys reviewed and stored safely, comfort controls set, any exposed credentials revoked and reissued, and your real details audited out of your personas and character notes.
Repeat the sitting twice a year. Safety is not a state you achieve but a light routine you keep, and twenty minutes is the honest price of it.
Protect the real things — account, keys, identity — and let the invented things flourish freely inside that protection. The boundary is not a restriction on the fiction; it is what makes the fiction safe to enjoy.
Do the twenty-minute setup today. Future incidents are decided by present habits far more often than by luck.
Extend the same audit to the rest of your AI usage once a year. Accounts, keys, and confidences accumulate across tools faster than habits do, and a single afternoon that reviews all of them is the cheapest strong security measure available to a private person.
