Frequently asked questions

No. Filtering is a judgment call made by a model, and it will sometimes hide something you wanted or let through something you did not. Report it when that happens and it helps us tune the filters.

Because of that, Lotus Pond is not a safety tool. Do not use it as a parental control or as protection against content that could genuinely harm someone. It is not built or tested for that.

Basic filters match as the page draws and cost nothing you would notice.

Smart filters wait on an answer from our servers, so an item nobody has been judged on before takes a moment the first time. Once an item has been judged against a filter, the result is cached and shared, so the next person to meet the same item does not pay for the same analysis again.

Not at all. Hiding happens in your browser, in your session. Nothing is reported, downranked, or removed on YouTube or Reddit, and creators are not notified or penalized.

Yes. That is the greenlist, which has its own page in the dashboard. Paste a link or type a name, and we look it up and show you what we found before anything is saved. A source on your greenlist is never filtered, by any of your filters.

One name can be several real sources. There is a YouTube channel called mkbhd, a subreddit about him, and a Reddit user with the same name, and they are three different things. Whichever of them we find are offered together, already ticked, so you untick the ones you did not mean.

What we can confirm differs by platform. A YouTube channel is checked against YouTube's own data API. Reddit refuses requests from our servers, so a subreddit or a Reddit user can only be checked when you have the extension installed, which asks Reddit from your own browser. When it could not be checked the card says so and gives you a link to look for yourself. Either way an entry works by recognising the names a post carries rather than by identifying an account, which is why the card shows you what it will recognise before you save it.

We do not sell personal information, and we run no advertising or analytics trackers. We train and test our own filtering models only on reports you send while training is switched on in your account settings, which is off until you switch it on and is offered only to adults. The AI providers we use may not train their models on what we send them.

OpenAI and Google. They are sent the item being judged and the filters it is being judged against, which means they see the wording of your own filters as well as the item. A model cannot decide whether something matches a filter it has not been shown.

They are never sent your username, display name, email address, account identifier, or payment details. They see items from a feed without being told whose feed it is. The one nuance: screening requests carry a routing key so the provider can reuse a cached copy of our instructions, which keeps the service fast and cheap. It is derived from your account by a one-way function and cannot be turned back into it. The Privacy Policy covers this in full.

Your account and most of what is tied to it, yes. Use Delete account in your account settings, or write to [email protected] and ask. You can also request a copy of your data before you go.

Deleting your account removes your sign-in, filters, settings, usage counts, consent records, stored pictures and billing customer (Stripe keeps its invoices). Records of past checks that are still linked to your account are deleted; older ones, which we have already unlinked, are kept without your account attached, and so are our usage and cost logs. We delete your support messages and misfilter reports with your account, at once, except any we must keep for a legal reason such as a payment dispute, which we keep unlinked from your account until the dispute ends. Messages you sent while signed out aren't linked to an account; ask us and we will delete them.

One thing happens on a timer whatever you do: after thirty days, we remove your account and your filters' wording from the record of each check itself, though some separate records can still connect it to you. The Privacy Policy spells out both.

From your dashboard, under subscription. Cancelling stops the next renewal and you keep access until the period you already paid for runs out.

Stripe. Card numbers go straight to them and never reach our servers.

Items the extension has checked and let through carry a small report control in the feed. Use it and pick the filters you expected to catch the item; that tells us far more than the link on its own. The contact page takes the same report if you would rather write to us.

When Lotus Pond cannot get an answer about an item, it holds the item back rather than showing it unchecked, and says on the page that it did. The usual causes are a signed-out session, a dropped connection, or reaching the day's filtering limit.

The notice carries a control that reveals what is being held, so you are never stuck with a blank feed. The status page says whether the problem is ours.

Both usually mean the same thing: you are signed out. Your filters live with your account, so sign in again from the dashboard and the extension picks them back up.

If you are signed in and nothing is being filtered, the platform may have changed its page layout, which stops the extension finding feed items until we ship an update. The status page will say if we already know. If filters still will not save, tell us. That is a bug, not something you are doing wrong.