The app creates an anonymous account for you the first time you open it, so your chats can sync and your daily message allowance can be counted. That account is a random identifier. It is not linked to your name, email, phone number, or device advertising identifiers. You can optionally add an email address so you can restore your history on another device. Adding an email is never required to use the app. The only other time we ask for an email address is the optional newsletter described under Email newsletters.

Some information never leaves your phone or browser. We do not receive it, and it is removed when you delete the app or clear the site data:

So your history survives a lost phone, the following is stored in our database and private file storage, under your account:

History Daily does not run its own AI models. When you send a message, your conversation with that figure, any image you attached, and any saved memories for that figure are sent to OpenRouter, which routes the request to the model provider that generates the reply. The same happens when the app extracts memories or builds a quiz from a conversation. When you tap Listen to have a reply read aloud, the text of that reply (which the server already generated) is sent to OpenRouter to be turned into speech by its speech provider; the audio is then kept only in your device's temporary cache, and we do not store audio on our servers. We do not control, and cannot promise, how those providers handle data once it reaches them; if that matters to you, review OpenRouter's policies before sharing anything sensitive. Do not send information you would not want processed by an outside AI service. Replies are generated text and can be incomplete, mistaken, or anachronistic. They are not professional advice.

Every message you send, every image you attach, and every reply the model produces is checked by an automated safety classifier (OpenAI's moderation service) before it is delivered or answered. The check looks for sexual content, sexual content involving minors, self-harm, threats, hate, and graphic violence. Most messages are never stored by the screening layer. When something crosses a safety threshold, we keep a copy of that message, the safety categories it triggered, and your account identifier so a person can review it. Messages that were flagged but allowed are deleted after 90 days. Blocked messages and the evidence around them are kept for up to one year for safety review. If a message of yours suggests you may be in crisis, the app shows crisis-support resources (988 and the Crisis Text Line in the US) instead of, or alongside, the figure's reply.

When "Let figures remember me" is on, the app asks an AI model to pull durable details out of your conversation and saves them so later chats with that figure feel continuous. This setting is on by default. Turning it off stops new memories from being saved and stops existing ones from being used in conversations, but it does not delete memories already saved. You can read every memory a figure holds, delete them one at a time, or clear all of a figure's memories, from that figure's profile page. Memory is not shared between figures.

Subscriptions are handled by RevenueCat and the app store or payment processor you buy through. They receive your account identifier so your purchase can be matched to your account. We receive only whether your subscription is active, when it renews or expires, and a link to manage it. We never see or store your card number.

We use PostHog, a product analytics service run by PostHog Inc. in the United States, to understand how the app is used and where it breaks. It receives events such as "quiz completed" or "chat message sent", with coarse details only: which historical figure you chatted with, your learner type, whether you have Premium, and similar. It also receives error reports — a technical stack trace of what went wrong, never the text of your messages. In the iPhone and iPad app it can record session recordings so we can see where things fail. Recordings only ever cover screens without your content — onboarding, the figure catalog, and settings — never chats, quizzes, study, or the Today feed; everything you type is masked, and recordings are kept for 30 days. There are no session recordings on the web app. If you have linked an email to your account, that email is attached to your analytics record so it can be matched to you. Analytics events never include the content of your conversations, quiz answers, memories, or attachments. Analytics are on by default; you can turn them off at any time with the "Help improve History Daily" switch in Profile, and nothing further is sent after you do.

We send a weekly history digest and occasional product news, but only to people who chose to subscribe and then confirmed that request from their email inbox. This is separate from the email you may add for account linking. A newsletter address is stored in our database and with our email provider, Resend, and is used to send those emails and nothing else; we do not track whether an email was opened or which links were clicked. Every email has a one-click unsubscribe link, and unsubscribing takes effect immediately. Whether you subscribe or unsubscribe never affects your access to the app.

There is no advertising SDK in this app. Product analytics and error reports go to PostHog, as described under Product analytics, and never include the content of your conversations. We do not sell your personal information, and we do not use your conversations to advertise to you. Our own server logs are written to record whether an operation succeeded or failed; they are not designed to contain the content of your messages. Read-aloud sends the reply's text to OpenRouter to be synthesised (see above) but never uploads your microphone or any other audio.

We keep your conversations, attachments, and memories until you delete them or ask us to. There is no automatic expiry, because a history you did not choose to lose is the point of the product. When you delete chat data, conversations, messages, and attached images are removed from our storage and are also blocked from being restored by another device that was offline at the time. Safety records described under Safety screening are the one exception: flagged content retained for review (90 days) and blocked-content evidence (up to one year) is kept separately from your chat data and is not deleted by the in-app delete buttons, because preservation and reporting duties can outlive an account deletion request.

Everything below is in Profile, under Privacy & Data:

Deleting your account does not cancel a paid subscription, because subscriptions are managed by the app store or payment provider rather than by us. Cancel it there first, or you may continue to be billed for a subscription attached to an account that no longer exists. On iOS this is in Settings, under your name, then Subscriptions.

If you would rather not use the in-app controls, or you want to know what is held about you, email the address below and we will action it.

History Daily is intended for people aged 13 and over, and is not directed to children under 13. We do not knowingly collect personal information from children under 13. If you believe a child under 13 has used the app and provided personal information, contact us at the address below and we will delete it. If you are under 18, please ask a parent or guardian to read this policy with you before using the app.

Data is stored with our infrastructure providers, Supabase and Vercel, product analytics is handled by PostHog as described above, and information is processed by the AI providers described above. This may involve transferring information to countries other than the one you live in. We use per-account access controls so that one account cannot read another's conversations, memories, or images, and attachments are stored in private storage rather than on the public internet. No online service can promise absolute security.

If we change how the app handles your information, we will update this page and the date shown at the top. Continuing to use History Daily after a change means the updated policy applies to you.

Questions about this policy, requests to see or delete your information, and privacy complaints can all be sent to chatwithhistorians@towb.in.