Beta: these terms are being reviewed by counsel and may change.
Legal
Privacy Policy
Last updated October 25, 2026
This policy explains what personal data XCap ("we") collects when you use XCap, why, who we share it with, and the choices you have. We are based in the United States. For account data we are the controller; for content and contacts you put into your workspace (for example leads) we act as your processor.
1. What we collect
- Account data: name, email, a hashed password (we can't read it), workspace name, role and email-verification status.
- Sign-up record: when you create an account with our sign-up or invite form, the version of the Terms of Service and Privacy Policy you accepted, when you accepted them, and when you confirmed you're 18 or older; if you used an invite code, which code.
- Free-trial checks: when a new account gets a free trial, we keep keyed hashes (one-way codes made with a secret key, which can't be turned back into the original) of the sign-up IP address, its network (the first three parts of an IPv4 address, or the first four blocks of an IPv6 address) and the email address, with the time. Before the email address is hashed, anything after a “+” in its name is dropped, and for Gmail addresses so are the dots, so variations of one inbox count as the same address. We use them only to stop repeat free trials (legitimate interests): the IP address and network count for 30 days, the email address for as long as its hash is kept. We check the email address only once you've confirmed it's yours. They are kept for 12 months, also after the account is deleted.
- Invite requests: while XCap is invite-only, if you ask for an invite we keep the email address, name, company and note you give us, when we invited you, and a keyed hash of the IP address the request came from (to stop abuse). We use them to decide who to invite and to send your invite (legitimate interests). We delete a request 180 days after we invite you, or 365 days after you made it if we don't, or sooner if you ask us or delete an account with that email address.
- Content: brand kits, scripts, captions, uploads, saved reference videos, format templates, products, and generated audio, images and video.
- Voice samples and face images (sensitive): recordings you upload to clone a voice and photos you upload to create an AI twin presenter, each with a consent record (name, statement, time, IP). See "Voice clones and AI twins" below.
- Connected platforms: when you connect YouTube, TikTok, Pinterest, LinkedIn or Meta (Facebook Pages/Instagram/Ads) we store encrypted access tokens, the account id and name, granted scopes, the posts we made for you (platform post id, link and status), and the statistics we sync for your own posts and ads.
- Billing: plan, credit balance and history. Card details are handled by Stripe, not stored by us. If you subscribe through the Shopify App Store, Shopify handles that charge (plan name, period and whether a charge is a test); we store the subscription status Shopify sends us, not your card.
- Results data: clicks on tracked links (with a daily-salted hash of IP and browser, not the raw IP), and leads you send us, where the lead's email is stored only as a hash.
- Connected stores: if you connect Shopify or WooCommerce, we store encrypted access tokens (or WooCommerce API keys), the shop domain, currency and time zone, product catalog fields we import, and attributed order records (order id, amounts, status, discount codes and tracking refs — never customer name, email, phone or address). See "Connected stores (Shopify and WooCommerce)" below.
- Technical data: session records, security logs, and rate-limit counters keyed by IP address.
- Public template gallery: anonymous counters for template pages (clicks, free script previews, sign-ups and first videos per template per day). These rows hold no IPs, cookies, user ids or workspace ids. Curated preview topics may be cached without your free-text input.
- Support: messages you send us.
2. How we use it
- To provide the Service: generate content, run jobs, post the videos you approve to the accounts you connect, attribute store orders you connect, and report results (contract).
- To secure accounts, prevent abuse and fraud (including repeat free trials), and enforce limits (legitimate interests).
- To bill you and keep required financial records (legal obligation).
- To send account emails such as verification, password reset and job notices, and, if you opt in, product updates you can unsubscribe from (contract / consent). Four emails in Settings → Notifications work the other way: approval needed (to a brand's approvers, or its owners when it has none, when a piece is waiting for approval), video ready (to the person who started a video, when it finishes and they're not on its page), the low-credit email (sent once, when your credits no longer cover a video) and the weekly summary (last week's results and next week's plan) are on by default, and an owner or admin can switch each one off there.
- Aggregated statistics. We may publish aggregated, de-identified usage statistics on public pages — for example how many businesses used a video format in a week. Counts are bucketed (for example "25+") and never shown for fewer than 10 businesses. We do not publish raw counts, workspace names or anything that identifies a business or person.
3. AI processing
Standard features run on models we host ourselves. Premium features, where offered, also run on models we host ourselves. We don't use your content to train generative AI models.
XCap Score. XCap Score predicts how a post will do. It learns from posts across workspaces, but only where a workspace has turned on "Help improve XCap Score" in Settings: the wording of hooks and scripts, posting details (platform, day, time and format) and results (views, clicks, leads, sales and A/B test winners). Only the trained model is shared between workspaces, never your posts or results. It is off until you turn it on, and you can turn it off again at any time. Changes apply from the next training run: a workspace that turns it off is left out of the next shared model, and older versions that learned from its posts are deleted then. Its own scores still tune to its own results either way. If an agency turns it off, that also covers its client brands. Data from connected Shopify and WooCommerce stores is never used, whatever the setting, including for a store's own scores. Numbers worked out from YouTube data (how fast public videos and topics are growing, and the views of your own YouTube posts) are used only while Google allows us to calculate them, and a model trained with them is deleted 30 days after it was trained, or at once if that permission ends. Each model records which kinds of data it learned from. We do this only with your consent, which you can withdraw in Settings, to make the Score more accurate for everyone who uses it.
Safety screening. To prevent abuse, an automated check looks at the photos you upload for AI twins or to animate, and at the images and videos XCap makes. It screens out minors, public figures and copyrighted or trademarked characters (likeness and abuse prevention). It only answers yes or no: it never tries to work out who someone is, and it keeps no face template.
4. Content check
XCap checks the text you send to make something (for example topics, prompts, scripts, captions, product details and agent settings) against its content rules: sexual content involving minors, explicit sexual content, nudity, hate, threats, graphic violence, self-harm instructions, illegal goods, scams and faking a real person, and, in pictures, videos and voices, public figures and copyrighted or trademarked characters. Unproven health claims and guaranteed money results are flagged, not refused. Fixed rules run first; when they aren't sure, or the text is long, a model we host ourselves reads it too.
- What we log. When the check refuses or flags something from your workspace (so does the safety screening above, with the prompt where there is one), we keep a log entry: the workspace and, when known, the person, where in XCap it happened, the categories, the text's length, a fingerprint (hash) of the text and its first 80 characters. Never the full text.
- How long. We delete the 80 characters after a year; the rest of the entry stays with the workspace's records. A child-safety case that was reported to the CyberTipline of the National Center for Missing & Exploited Children, or held for legal advice, keeps its 80 characters until a year after that review.
- Deleted accounts. When we must keep records for legal or safety reasons, an entry can outlive the account or workspace it is about: we keep it, with the account's email and the workspace's name, for as long as we must (its 80 characters follow the rule above). After that we delete the entry, or only the email and name if its workspace still exists.
- Who reviews it. XCap's platform admins, who sign in with a second step, review these entries, and each decision is recorded with the admin who made it. A refusal involving a minor goes to a separate child-safety queue, where an admin sees its 80 characters only by opening it, and each opening is recorded.
- Admin views are logged. Each time an admin looks at these entries we record which admin, which workspace (or all of them) and when, at most once an hour per admin, workspace and page.
- Suspensions. If we suspend a workspace, we record when, which admin and the reason, and email the reason to its owners. Lifting a suspension is recorded the same way.
5. Voice clones and AI twins (sensitive data)
A voice sample or a face image can identify a person and, in some places, counts as biometric or special-category data (for example under the Illinois Biometric Information Privacy Act (BIPA), Texas and Washington biometric laws, and GDPR / UK GDPR Art. 9). We treat them as sensitive:
- Explicit consent. We only create a voice clone or AI twin after the person heard or shown gives explicit written consent in the app. We record who consented, the exact statement, the time and the IP address.
- Purpose limitation. Samples and images are used only to generate the voiceovers and presenter videos you ask for in your workspace, and to run the safety screening described under AI processing (screening out minors, public figures and protected characters to prevent abuse). We don't use them to identify people, for surveillance or advertising profiles, or to train general AI models.
- No sale or sharing. We never sell, lease, trade or otherwise profit from voice samples, face images, or any voice or face model or template derived from them. They go only to the processors needed to generate your content, which are our own hosted models.
- Retention. We keep them until you delete the voice or presenter, delete your account, or the person withdraws consent — and no longer than 3 years after you last used that voice or presenter, or sooner where the law requires.
- Deletion. Deleting a voice or presenter (or your account) removes the sample or image from storage and unregisters the cloned voice at the voice engine. We keep a record of the consent for 3 years as proof that it was given: the person's name, when they consented, and fingerprints (hashes) of the consent statement and IP address, never the sample or image. The person whose voice or face it is can also ask us directly at [email protected].
6. Connected platforms and YouTube API Services
XCap uses the YouTube API Services. By connecting YouTube you agree to the YouTube Terms of Service, and Google's use of your data is covered by the Google Privacy Policy.
XCap's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.
- With your permission we access your channel id and name and per-video statistics such as views, average percent watched, likes, comments and shares, to show results for videos you made with XCap.
- With the upload permission, XCap uploads a video to your channel only when you approve that post (or turn on auto-approve for a Series), with the title and description shown in Publish. Until Google approves XCap's app, YouTube keeps these uploads private. We don't edit or delete your videos.
- We refresh this data regularly and delete YouTube API data we haven't refreshed within 30 days.
- You can revoke our access at any time from Settings → Connections or at Google security settings. When access is revoked we delete the YouTube data we stored — immediately when you disconnect in the app, and within 7 days if you revoke at Google.
- Posting for you. XCap uploads and posts videos to the YouTube, TikTok, Instagram, Facebook Page, LinkedIn and Pinterest accounts you connect only when you approve a post (or turn on auto-approve for a Series), at the time shown in Publish, with the caption shown there. You can cancel a scheduled post until it goes out. While a platform is still reviewing XCap's app, posts there may go out as private (only you can see them) or wait until the review passes; the app tells you when that's the case.
- TikTok, Pinterest, LinkedIn and Meta work the same way: disconnecting revokes our token where the platform offers a revoke API (Pinterest and LinkedIn don't, so we delete the token) and deletes the statistics we synced from that platform. Statistics you entered yourself are kept.
- Paid ads run in your own Meta ad account. If you delete your account, campaigns and ads stay in your ad account; we delete our copies and metrics.
- Platform data is never sold, never used for advertising, and only shared with the platform itself to do what you asked.
7. The XCap app on your devices
- App files on your device. When you use XCap in a browser or install it as an app, your browser keeps a small cache of its icons and a built-in offline page, so it opens quickly and can tell you when you're offline. Neither holds personal data; both stay on your device and aren't sent to us.
- Offline review. If you save videos to review offline, your device keeps the review queue, light preview copies of those videos and any decisions waiting to send. They expire 72 hours after your device was last online, and are wiped when you sign out or someone else signs in on that device.
8. Connected stores (Shopify and WooCommerce)
If you connect a Shopify or WooCommerce store, we use that connection only to run XCap for that workspace: import products you choose, write ads, attach real discount codes you confirm, and attribute paid orders to the post that brought them. We never use store data to train any model, including XCap Score.
- What we access. Shopify: products (title, handle, photos, price, description, publication status) and orders at Level 1 protected data (order id, times, financial status, currency, amounts, discount codes, landing-page and cart refs, and when available a customer-journey summary of landing refs). We do not request or store customer name, email, phone or address. WooCommerce: the same product and order fields after we strip customer personal data from webhook payloads on arrival.
- What we store. Encrypted tokens or API keys; shop domain and settings; catalog rows; attributed paid orders (net of refunds) as leads with no contact details; orders we could not trace to a post (order id, amounts and status, no contact details), which a daily cleanup deletes once they are 48 hours old (a large backlog can take more than one daily run), leaving only a daily count of them (no amounts or ids); the store's webhook message that brought each order, with the same order fields, until a daily cleanup 30 days after we processed it, or 90 days after it arrived if we couldn't; offer and discount-code records you confirm. Untraced orders never appear in revenue totals.
- Retention and deletion. While the store is connected we keep this to provide the Service. When you uninstall the Shopify app, we delete its access tokens at once; Shopify then sends a shop/redact request (about 48 hours later), and when it arrives we delete the Merchant Data — events, codes, offers, orders, API-synced products, attributed leads, product-photo copies and the order data in related job rows — keeping only the store's record (its Shopify address and id and billing fields, without tokens or settings). A customers/redact request deletes the orders it names. Disconnecting a store in XCap, Shopify or WooCommerce, deletes its tokens or API keys at once but not the rest of its data, and WooCommerce has no deletion request we can act on: that data is deleted when you delete your XCap account (see "How to delete your data") or ask us at [email protected].
- Trial-abuse hash. To stop repeat free trials and sandbox grants on the same Shopify shop after deletion, we keep a keyed hash of the shop's id (not the domain, and not any customer). It cannot be turned back into the shop id, and it is used only for that check.
- Sub-processors. Store data is hosted with the providers in the table below (hosting, database and media storage). Shopify Billing, when you subscribe in the App Store, is handled by Shopify under Shopify's terms; WooCommerce stays on your own site.
9. Sub-processors
We share data only with the service providers needed to run XCap:
| Provider | Purpose | Data |
|---|---|---|
| Stripe | Payments, invoices and subscription management | Billing contact, payment details (held by Stripe) |
| Shopify | App Store billing when you subscribe in Shopify; receiving the product and order fields you connect | Shop domain, plan and charge status (held by Shopify); product and Level-1 order fields we request (no customer name, email, phone or address) |
| Pexels | Stock footage search | Search terms (no personal data) |
| Google — YouTube API Services | Trend research; when you connect YouTube, your own channel's analytics and uploading the videos you approve | Channel id and name, per-video statistics, OAuth tokens, videos and titles you post |
| Google Trends | Trend research | Search terms (no personal data) |
| Meta (Facebook, Instagram, Marketing API, Ad Library) | Posting the videos you approve to Instagram and Facebook Pages, their insights, ads in your own ad account, ad research, public Instagram posts for the hashtags you follow on your Radar, and details of public Instagram links you save | Page/Instagram ids and names, post and ad metrics, OAuth tokens, videos and captions you post, ad creatives you launch, hashtags you follow, links you save |
| TikTok | Posting the videos you approve and your own video statistics when you connect TikTok | Account id and name, video statistics, OAuth tokens, videos and captions you post |
| Posting the pins you approve and pin statistics when you connect Pinterest | Account id and name, boards, pin statistics, OAuth tokens, pins you post | |
| Posting the videos you approve when you connect LinkedIn | Member id and name, OAuth tokens, videos and post text you post | |
| Buffer | Scheduling and publishing posts when you connect it | Posts, media and captions you schedule |
| Email provider | Sending account and notification emails | Name, email address, message content |
| Cloudflare R2 | Storing media files | Uploaded and generated media, voice samples, presenter images |
| Hosting provider | Running the app, database and job workers | All Service data, encrypted in transit |
We'll update this list before adding a new sub-processor.
11. How long we keep data
- Account and workspace data: while your account is active.
- When you delete your account, workspaces you own alone and their content — media files, voice samples, AI twin images, platform tokens (revoked at the platform where possible) and synced statistics — are deleted immediately, and so are your sign-up record and the support messages you sent us. Deleted data can remain in encrypted backups for a limited time before they are overwritten. We keep a log entry with counts only (no personal data) as a record of the deletion; when one of our admins deletes an account, a second entry records which admin, the reason they gave and the ids of the deleted account and workspaces. For 3 years we also keep: the consent record of each voice and AI twin (see "Voice clones and AI twins"); and, with account and workspace ids replaced by keyed hashes, a record of the deletion request, of blocks and safety escalations involving minors (their category, where and when, never the content), and of copyright notices. When we must keep records for legal or safety reasons, we keep the ones we need, such as a content-check log entry with the account's email, for as long as we must (see "Content check").
- YouTube API data: refreshed or deleted within 30 days, and deleted when you revoke access.
- XCap Score models trained with numbers worked out from YouTube data: deleted 30 days after they were trained, or at once if Google's permission ends.
- Connected stores: Shopify store data is deleted when Shopify's shop/redact request arrives (about 48 hours after you uninstall the app); WooCommerce store data, and that of a store only disconnected in XCap, until you delete your account or ask us; orders we could not trace to a post are deleted by a daily cleanup once they are 48 hours old (a large backlog can take more than one daily run), leaving only a daily count (no amounts or ids); the store's webhook messages (one about an order holds the same order fields) are deleted by a daily cleanup 30 days after we processed them, or 90 days after they arrived if we couldn't; the trial-abuse shop hash is kept after deletion only to prevent a second trial on the same shop.
- Voice samples and face images: see "Voice clones and AI twins" above.
- Content-check log entries: their 80 characters of text for a year (a child-safety case reported to the CyberTipline or held for legal advice: until a year after that review); the rest stays with the workspace's records and is deleted with them, unless we must keep it for legal or safety reasons. See "Content check" above.
- Billing records: as long as tax and accounting law requires (typically 7 years).
- Offline review data on your device: 72 hours after your device was last online, or sooner when you sign out.
- Free-trial check hashes: 12 months from sign-up, including after the account is deleted.
- Invite requests: 180 days after we invite you, or 365 days after the request if we never do (sooner if you ask us or delete an account with that email address).
- Rate-limit counters (keyed by IP address) reset when their time window ends and are cleared from the database from time to time. Security logs are kept only as long as needed to protect accounts and the Service.
- Password-reset links expire after 1 hour; email-verification links after 48 hours.
12. Your rights (GDPR, UK GDPR, CCPA/CPRA)
Depending on where you live, you may have the right to access, correct, delete, or receive a copy of your personal data, to object to or restrict processing, and to withdraw consent. California residents have the right to know, delete, correct, and opt out of sale or sharing (we don't sell or share), and not to be discriminated against for using these rights.
- Do it yourself: Settings → Account lets you export your data as JSON and delete your account. If Settings can't delete it, write to [email protected] and we'll take care of it.
- Or ask us: email [email protected]. We'll verify your identity and respond within 30 days (45 for California requests).
- If you're in the EEA or UK, you can complain to your local data protection authority.
If you're a lead or contact of one of our customers, please contact that business first; we'll help them respond.
13. How to delete your data
- Delete your account: Settings → Account → Delete account. If Settings can't delete it, write to [email protected] and we'll take care of it.
- Disconnect one platform: Settings → Connections.
- Disconnect a Shopify or WooCommerce store: Settings → Connections, or uninstall the XCap app in Shopify admin / remove the WooCommerce keys. Disconnecting deletes the store's tokens or keys at once. Shopify store data is deleted when Shopify's shop/redact request arrives, about 48 hours after you uninstall the app; for WooCommerce data, delete your account or email us.
- Removed XCap from Facebook or Instagram? Meta sends us a deletion request and gives you a confirmation code you can check at /data-deletion, which also lists every option.
- Or email [email protected].
14. International transfers
We and our providers process data in the United States and other countries. Where required, we rely on Standard Contractual Clauses or other lawful transfer mechanisms.
15. Security
We hash passwords with scrypt, encrypt stored API secrets, use HTTPS, limit login attempts, and restrict staff access. No system is perfectly secure; we'll notify you of a breach affecting your data as the law requires.
17. Children
XCap isn't for anyone under 18, and we don't knowingly collect their data.
18. Changes and contact
We'll post changes here and email you about material ones. Questions: XCap, 6408 E Virginia Ave, Scottsdale, AZ 85257 — [email protected].