TAEONA

Privacy policy

Who we are

TAEONA is a product of Taeona, Inc., 742 S Hill St, STE 903, Los Angeles, CA 90014. Taeona, Inc. is the data controller for the information described here. Privacy questions and requests go to hello@taeona.com — see support.

This policy is readable without an account and covers visitors, waitlist subscribers, customers, the people in a customer's workspace, and anyone whose information reaches us through the product. Our customers are in the United States, Canada, the United Kingdom and the European Union.

Our role, and whose policy applies

Two different things happen here and the law treats them differently.

• For running our own business — your account, sign-in, support, security and, once billing is on, payment — we decide how the information is used. Taeona, Inc. is the controller.
• For the material inside a workspace — a customer's brand, catalogue, prompts, drafts, published posts and connected accounts — we act on that customer's instructions. If you are an employee, contractor or agency client of a TAEONA customer, that customer's own privacy notice applies to you as well as this one, and a request about their content may be one we pass to them while helping as the law requires.
• A platform you connect — Meta, TikTok, X and the rest — handles what it receives under its own policy, independently of us.

An integration described here is offered only where the feature is available and you have authorised it. Reading about one does not mean we hold that data about you: if you never connect a social account, nothing arrives through that account's authorisation.

What the product does with your data

This section describes the software's actual behaviour, written from the code and accurate as of 2026-10-13.

• Your account. Your email address; your name and profile picture if you sign in with Google; your phone number if you sign in by SMS. There is no password: sign-in is a one-time code, or Google.
• Sign-in records. Each session stores the IP address and browser user-agent it was created from, and each one-time code stores the address or number it was sent to and the IP address that asked for it. Both exist to stop other people's accounts being attacked and to stop paid SMS being pumped. Sessions and codes are stored hashed, never in a form that could be replayed.
• Invitations and the waitlist. The beta is invite-only, so we keep the addresses invited, who invited them and when. If you joined the waitlist on our website we keep the name, email, optional brand website and which form you used.
• Your website and catalogue. When you supply a website address, TAEONA reads public pages and, for Shopify stores, the public product catalogue, and stores the product photography it finds so creative can be generated from it.
• The other photographs on your website, including ones with people in them. TAEONA also stores the pictures on those pages that belong to no product — the shopfront, the room, the counter, a hero banner, your team. A business with no catalogue has no other photography of itself, and a post has to be made of something. People appear in these: staff, and whoever your own site shows in the room. A few from each crawl (six by default) are sent to OpenAI to be described in words, and the description is stored with the picture so the assistant can find the right one later. They are kept for the life of the workspace, filed under no product, marked as having come from your site rather than from you, and any of them can be deleted on its own.
• Your business's public Google listing. If your site carries little photography, we look the business up by name and address in Google's Places API and store photographs from the listing — some of which were uploaded by customers rather than by you — plus a few short review excerpts as background notes for writing captions.
• A public social profile you name. If you ask the assistant to read a brand's own public Instagram profile, we fetch it through a third-party reader and store what it returns alongside your brand notes.
• What you upload. Images and logos you add yourself.
• Content you create. Your conversations with the assistant, captions, image and video briefs, the generated images and video, and the schedule they are planned against. Automated quality checks on generated frames are also stored.
• Formulas you list in the marketplace. TAEONA has a marketplace where a workspace can list a creative formula it has written and other workspaces can license it. Listing one is your choice and nothing is listed unless you list it. When you do, a shop window is created that every other signed-in TAEONA workspace can see and search: a seller name, the title, the pitch, the category, your tags, your “what’s included” lines, the price, how many times it has been licensed, and the pictures in the window. The seller name is your own name on your account by default, not your brand — you can change it as you list, and you should if you would rather not be named. The title and the pitch are taken from the formula’s own name and premise unless you rewrite them, so words you wrote for yourself become words other people read. If you pick no picture we make a cover from the formula and publish that. Separately, and never shown to anybody browsing, we keep a frozen copy of the formula’s full wording — that is the thing being licensed, and it reaches another workspace only when they license it.
• The pictures on a listing are public. A listing’s cover and window images are served without a sign-in, so anyone with the link can open them and search engines may index them. That is what makes a shop window a shop window. Taking the listing down closes those pictures immediately.
• Sound in generated video. Films generated by TAEONA carry an audio track, and it is generated with the picture by the same model — the room's own sounds, the action's sounds, and, when the brief calls for a spoken line, a voice reading a script this product wrote. No separate speech service is involved, and when a film has no script we instruct the model that nobody speaks in it. Where one film is rendered in several parts, a short sample of the first part's audio is sent back to the same provider with the later parts so the voice stays the same across the cut.
• Connected accounts. When you connect a social or commerce account, TAEONA stores an access token, the account's name or handle, its follower count where the platform returns one, and the list of destinations (pages, channels, boards, locations, shops) you can post to. Tokens are encrypted at rest with AES-256-GCM and are never shown in the interface, written to logs, or included in error messages.
• Publication records. For each post published, the platform's own identifier and public URL, who approved it, and when.
• Publishing-permission records. Whether this workspace approves each post or has turned automatic publishing on, and for each change: the email address of the person who made it, the time, and which version of the agreement they accepted. No IP address and no browser details are stored with it.
• Credits, and payments once billing is switched on. Your workspace's credit balance and the ledger of what was granted and spent. When billing is enabled, a record of each purchase: which pack, the amount, the currency, the time, and the identifiers Stripe gives us. Card numbers are entered on Stripe's own payment page and never reach us.
• Feedback you send us. The message, the screen you were on, your browser user-agent, any screenshots you attach, recent errors from your browser, and a snapshot of the server's recent log lines so the problem can be reproduced.
• Server logs. Ordinary operational logging, which can include email addresses, workspace identifiers and, for a post that was published, the platform's id for it or its link.

What it does not collect

• No customer or order data from your shop. The Shopify integration requests product, file and content permissions only. It does not request access to orders, customers, addresses, email addresses or payments.
• No card details. When billing is switched on, payment details are entered on Stripe's hosted checkout page. We never see or store them.
• No advertising or analytics tracking in the product. One essential cookie keeps you signed in.
• No voice recognition, and nobody's real voice. TAEONA does not transcribe audio, does not accept audio you upload, and does not clone or imitate a real person's voice. The only voice it produces is one the video model performs from a written script.

Who decides what gets published, and when

Connecting an account does not authorise a post. How posts reach your accounts is a setting on your workspace, and it is yours to change:

• Approve each post — the default. Nothing goes to a connected account until someone in your workspace approves that specific post. The approval covers that exact content: if the post is edited afterwards the approval stops applying and we ask again. An approved post goes out at the time it was scheduled for.
• Automatic publishing — only if you turn it on. It is off until the workspace owner turns it on, which is a separate step with its own wording: we record who turned it on, when, and the exact wording they were shown. You can turn it off again at any time, and turning it off withdraws the posts already queued that nobody approved individually. Today this setting posts to Instagram and nowhere else. Meta’s rules would also permit it for a Facebook Page, but we do not plan posts for Pages, so a Page receives only what someone approves post by post.
• Some platforms always ask. Under their own developer rules, a person has to act on each post for X, TikTok, YouTube, LinkedIn, Pinterest and Google Business Profile, so those accounts stay approve-each even while automatic publishing is on, and the app says so next to each of them.

Why we process it, and on what basis

• To provide the service you asked for — reading your site, generating creative, scheduling it, and publishing the posts you have authorised. Basis: performance of a contract.
• To keep the service working and secure — error logs, rate limiting, and detecting abuse. Basis: legitimate interests.
• To take payment, once billing is switched on. Basis: performance of a contract, and legal obligation for the records kept afterwards.
• To meet legal obligations — retaining records of what was published on your behalf and of money taken. Basis: legal obligation and legitimate interests.

We do not sell your data, and we do not use your brand assets or generated content to train our own models.

How long we keep it

• Access tokens — deleted when you disconnect the account. For Facebook and Instagram, also when you remove TAEONA in their settings or they send us your deletion request; for Shopify, when Shopify tells us the shop's data is to be removed after the app is uninstalled. X, TikTok, Google, LinkedIn and Pinterest do not tell us when you revoke access in their settings: the token stops working at once, and our encrypted copy is deleted when you disconnect it here or delete the workspace.
• Your workspace (brand kit, catalogue, generated media, calendar, conversations) — for as long as your account is open, then deleted within 30 days.
• Publication records — the record of a post that went out: the platform, its identifier there, and who approved it. These live WITH the post and the workspace, and are deleted with either — delete the post and its publication record goes too; close the account and they go with the workspace, within 30 days. While they exist we keep them up to two years, because they are what a platform asks for after the fact. Removing TAEONA in your Facebook or Instagram settings, or a deletion request for that account — through Facebook or Instagram, or by email — removes that account's id, the post ids and the links from these records straight away; what is left says only that a post went out, when, and who approved it. An in-app Disconnect leaves them as they are. The separate log of who switched automatic publishing on or off does outlive the account on purpose; it is described below.
• Sign-in and invitation records (invitations, one-time-code records, waitlist entries) — kept for twelve months. These are keyed to an email address rather than to a workspace, so they outlive a deleted account.
• Feedback you sent us, and its screenshots — kept after account deletion, because a bug report is addressed to us and has to survive the workspace it came from. Kept for two years.
• Billing records — once billing is switched on, the record of a payment is kept after the workspace is deleted, because tax and accounting rules require it and because a refund can arrive later. It holds the workspace id, the amount, the time and Stripe's identifiers; it holds no content. Kept for seven years, the period United States tax and accounting rules require.
• Publishing-permission records — kept after the workspace is deleted. One record is kept deliberately: the log of when a workspace turned automatic publishing on or off, and the email address of the person who did. It is the evidence that posting was authorised, it is the record a platform asks for after the fact, and it would be worthless if closing the account erased it. It holds no post content.
• Marketplace listings — a listing is taken down, never deleted. Taking it down removes it from the shop, closes its public pictures and stops anyone else licensing it; the row is kept, including the seller name and the frozen copy of the formula, because the licences already issued from it have to stay answerable — a licence whose listing has vanished cannot say what was licensed. Kept for two years after the last licence issued from it, or two years after it was taken down if none ever was.
• Marketplace licences — the record that one workspace licensed a formula from another: the listing, the two workspaces, the amounts and the time. It holds no content. Kept for seven years, for the same tax and accounting reasons as billing records. A buyer’s own copy of a formula lives in their workspace and is deleted with it.
• Deletion records — when posts or a workspace are deleted we keep a line saying what was removed and when, with identifiers and counts only and no content, so we can answer "where did this go". The same goes for a deletion request that reaches us from Facebook or Instagram: a confirmation code derived from it, when it arrived and what was done, so the status page can answer it.
• Logs — application logs are kept for 30 days. Our hosting provider keeps its own request logs under its standard retention, which we do not control.
• Backups — our database provider takes its own backups on its own schedule, which we do not set and cannot shorten. A record you deleted stays in a backup until that cycle ages it out, and a backup is never restored into normal use without the deletions being applied again. Where the law or a platform imposes a shorter deletion deadline than the backup cycle, the shorter deadline governs what we do with live systems and we do not treat the backup as a reason to miss it.

Deleting your data

Disconnect an account in the app to revoke its access and delete its credentials — any post already queued for it is cancelled at the same time. Delete account removes your user record, the workspaces where you are the only member, and the files in them; the exceptions are listed above. Facebook and Instagram users can also request deletion through Meta; see data deletion.

Marketplace listings are handled separately, and one part of this cannot be undone. Deleting a formula takes its listing down with it, wiping your workspace takes your listings down, and closing your account takes them down too. That removes them from the shop and closes their pictures. It does not reach the copies other workspaces have already licensed. Those were delivered to them, they keep them, and we have no way to take them back — which is worth knowing before you list something. Where a listing has already been licensed, the listing row and the licence records are kept as described above. To have a listing taken down for any other reason, report it from the listing itself, or write to hello@taeona.com.

Who else processes it

We use the following processors. Each receives only what it needs to do its job.

• Supabase — the database and the file storage that hold your workspace, including the encrypted tokens and the generated media.
• Railway — hosting for the application and its server logs.
• OpenAI — generating text and images, describing photographs, and web research for trends. Receives your brand information, your product photography and the briefs written from them. It also runs the marketplace check: every formula submitted to the marketplace is read automatically before it goes live, and the whole of its wording — premise, audience, emotion, setting, plot, action, hooks, elements, exclusions, cast and video brief — is sent for that check along with the listing’s title, pitch and tags. A listing put back up after being taken down is checked again from the formula as it stands now, because you can edit it while the listing is down — unless that formula is no longer in your workspace, in which case the frozen copy goes back up as it was, on the check it passed when it was first published.
• Atlas Cloud (Seedance) — generating video, and the sound and any spoken line in it. Receives the product photograph, the shot brief for that film, the spoken script when there is one, and — for a film rendered in several parts — a short sample of the audio it returned for the first part, so the voice does not change halfway through.
• Resend — email delivery: sign-in codes, beta invitations, waitlist confirmations, and the internal alert that tells our team you sent feedback (which carries your address and an excerpt of your message).
• Twilio — sending sign-in codes by SMS, where phone sign-in is offered. Receives the phone number.
• Google — three separate things: Google sign-in, if you use it, which returns your Google account id, email, name and profile picture; the Places API, which receives your business's name and address and returns the listing, its photographs and review excerpts; and Google Fonts, from which the sign-in page and the app load typefaces, so your browser's IP address and user-agent reach Google when a page loads.
• Apify — reading public web data: trend research, and the public social profile you asked us to look at. Receives the handles and URLs you supply.
• Firecrawl — reading public web pages during a crawl. Receives the URLs you supply.
• Cloudflare — DNS for our domains and hosting for our marketing website, including the waitlist form.
• Stripe — payment processing, when billing is enabled. Stripe collects your payment details directly on its own hosted page and tells us only the outcome, an amount, and identifiers. Until billing is switched on, no payment data exists at all.
• The platforms you connect (Meta, Google, TikTok, LinkedIn, Pinterest, X, Shopify) — receive the content you have authorised for publishing, and nothing else.

Each of these processes data on our behalf under its own standard service and data-processing terms. The marketplace is not an exception to any of this: what is offered there is a formula you chose to list, on terms you set, and nothing about your workspace, your catalogue, your customers or your published posts is offered with it. We do not sell data, we do not use it for advertising, and we do not share it with anyone outside this list.

We do not sell personal information, and we do not share it for cross-context behavioural advertising or targeted advertising. There is no advertising pixel, no analytics product and no data broker anywhere in this product. We disclose information to someone else only on your instruction, to the processors above, or where the law compels us.

AI processing, and the limits on it

Three providers above are AI providers, and this is what reaches each of them:

• OpenAI receives your brand information, your product and site photography, the briefs and captions written from them, and the wording of any formula submitted to the marketplace. It is used through OpenAI's business API, not a consumer chat product.
• Atlas Cloud (Seedance) receives the product photograph, the shot brief, the spoken script where there is one, and a short audio sample of an earlier part when one film is rendered in several pieces.
• Apify and Firecrawl receive the public URLs and handles you ask us to read. They are not model providers; they fetch public pages.

Both AI providers process in the United States. We do not train or fine-tune any general-purpose model on your content or on data from a platform you connected, and we do not authorise a provider to use it for their own model training. We give a provider platform-origin data only where the source platform permits it — your permission does not override a platform's prohibition. In particular, we do not use X content or X API data to train or fine-tune any model. Providers may hold an input briefly for abuse monitoring under their own business terms; that is their retention, not a licence for us or them to learn from your brand.

What the product remembers about a product. TAEONA writes down, in words, how a product works — which way a dial turns, what a label says — by reading the photographs you supplied and sending crops of them to OpenAI to be described. Those notes are stored with that product, in that workspace, and are used only to brief later content for that same workspace. They are not pooled across customers, nothing one workspace writes reaches another, and no social metric or audience data is sent to a model. The notes are deleted with the product and with the workspace.

We do not use the product to make solely automated decisions about anybody's eligibility for employment, credit, housing or anything else of similar significance, and we do not use connected-platform data to build profiles of people, infer sensitive characteristics, conduct surveillance, or target advertising in a way a platform forbids.

Where it is processed

Taeona, Inc. is a United States company and your data is stored and processed in the United States. That is the whole list: we run no infrastructure elsewhere, and the processors above serve us from the United States. If you are outside the United States, using TAEONA means your data is transferred there, and United States law gives it different protection from the law where you live.

Where a transfer out of the EEA, the United Kingdom or Switzerland needs a safeguard, the mechanism is the one in each provider's own data processing agreement — in practice the European Commission's standard contractual clauses with the UK addendum where it applies. We have not certified under the EU–US Data Privacy Framework and do not claim to have. Ask us at hello@taeona.com and we will tell you which mechanism covers a particular provider.

How it is protected

These are measures that are actually in place, not aspirations:

• Everything travels over HTTPS; the application refuses plain HTTP.
• Social access tokens are encrypted at rest with AES-256-GCM, are never rendered in the interface, never written to a log, and never included in an error message.
• There is no password to steal. Sign-in is a one-time code or Google. Sessions and one-time codes are stored only as hashes, so a copy of the table cannot be replayed, and a code is capped at a fixed number of attempts and then burned.
• Each workspace's data is kept to that workspace, except where a feature shares it by design — a marketplace listing you publish and the licences issued from it — and in the recent server-log lines attached to a feedback report. System tasks such as token refresh, publishing sweeps and platform callbacks look records up across workspaces and act only inside the workspace each record belongs to. Access to production is limited to the people who operate the service.
• Deleting a connected account deletes its credentials in the same step and cancels whatever was queued for it, so a disconnect cannot leave a usable token behind.

We do not hold SOC 2 or ISO certification and do not claim either, and no system is perfectly secure. If we have a security incident we will handle and notify it as the law and our platform obligations require. Report a concern to hello@taeona.com — please do not put a credential in the message.

Your rights, and how to use them

Depending on where you live you may have the right to access, correct, delete, export or restrict the processing of your personal data, and to object to processing based on legitimate interests. Most of these you can exercise yourself in the app — Disconnect for a single account, Delete account for everything. For anything else, write to hello@taeona.com and we will respond within 30 days. You may also complain to your local data protection authority.

Put Privacy Request or Data Deletion Request in the subject, and say which account email, workspace or connected handle you mean and what you want done. You do not need a TAEONA account to ask — if your picture or details reached us inside a customer's content, write to the same address. We verify a request proportionately and ask only for what is needed to establish who you are or what authority you have. Never send a password or an access token. Pausing publishing, disconnecting an account and asking for deletion are three separate things; ask for all three if that is what you want. Full instructions are on the data deletion page.

Deleting what we hold does not pull down a post already published on a platform, or a copy someone else took. Manage published posts on the platform itself, where you can also exercise your rights against that platform directly.

If you are in the EEA or the United Kingdom

Our legal bases are listed under “Why we process it” above: contract for providing what you asked for, legitimate interests for keeping the service working and secure, consent where the law requires it, and legal obligation for the records we must keep. If you are a person inside a corporate customer, our contract with that company is not automatically the contractual basis for processing your personal information. You can object to processing we base on legitimate interests, and to direct marketing at any time; withdrawing consent does not undo processing that was lawful before you withdrew it. Information a feature needs is necessary for that feature — you can decline, and the feature will not work. You may complain to your supervisory authority, including the UK Information Commissioner's Office. We have not appointed an Article 27 representative or a data protection officer, because on our assessment neither is required of us at our current scale and processing; if that changes we will appoint one and publish the contact here.

If you are in Canada

You may ask for access to and correction of your information, ask how we handle it, withdraw consent subject to legal and contractual limits, and complain to us or to the Office of the Privacy Commissioner of Canada or your provincial authority. Write to hello@taeona.com, which is also our privacy contact for this purpose. Your information is processed in the United States, as described above, and we will tell you what withdrawing a particular consent would cost you in function before you decide.

If you are in California or another United States state

Where a state privacy law gives you rights, use the same address and we will honour them. So that the position is unambiguous rather than buried: in the preceding twelve months we collected the categories of information listed under “What the product does with your data” above, from the sources, for the purposes, and shared with the recipients listed in this policy, and we kept them for the periods under “How long we keep it”. We have not sold personal information, and we have not shared it for cross-context behavioural advertising, so there is no opt-out to offer and no Global Privacy Control signal to honour — if that ever changes we will say so here and build the opt-out before, not after. We do not use the product for profiling that produces legal or similarly significant effects. If we refuse a request you may appeal by replying to our answer, and we will respond to the appeal in writing. This product is sold to businesses, and much of what it holds is business-contact and customer-content information rather than consumer information; we apply these rights to whatever is in scope rather than arguing about the boundary.

Cookies

We set one essential cookie, which keeps you signed in. There is no advertising or analytics tracking in the product, no analytics product of any kind, and no third-party tracking pixels. Because the only cookie we set is the one that signs you in, there is nothing to consent to and no consent banner.

One third-party request does leave your browser, and it is not a tracker: the sign-in page and the app load typefaces from Google Fonts, so your IP address and browser user-agent reach Google when a page loads. It sets no cookie for us and we receive nothing back from it. Our marketing website is served by Cloudflare, which keeps its own request logs to serve and protect the site. If we ever add anything that is not essential, we will ask for consent before switching it on and give you a way to change your answer, and declining optional tracking will never break signing in.

Children

TAEONA is a business tool and is not intended for anyone under 18, nor directed at children. We do not knowingly collect personal information from a child under 13. If you think a child has given us information, write to hello@taeona.com and we will look into it and act. If content you make includes anyone's image, voice or details — a minor's included — you must have the rights and the lawful basis for it.

Other services, and changes to this policy

When you connect or publish to another platform, that platform's own terms and privacy policy govern what it does with what it receives — Meta, TikTok and X publish theirs, and the version that applies to you may depend on where you are.

If this policy changes materially we will tell you by email and in the app before the change takes effect, and where a new use needs consent we will ask for it before that use begins. Updating the date on this page does not by itself authorise a new use of information, widen an automated action, or override a platform's restrictions — and it does not retroactively establish that notice was given before it actually was.

Contact

Taeona, Inc., 742 S Hill St, STE 903, Los Angeles, CA 90014, United States. Privacy and support: hello@taeona.com. See support, terms of service and data deletion.