Privacy policy
Last updated 6 October 2026
CastGenX is a podcast production studio. Your shows, scripts, audio and recordings are yours: we use them only to produce and publish your episodes and never to train AI models. This policy explains what we collect, why, who helps us process it and the choices you have.
1. Who is responsible
[Company legal name — OWNER], [OWNER: registered postal address], is the controller of the personal data described here. Privacy contact: support@castgenx.com. Data protection officer: [OWNER: DPO or privacy lead, if appointed]. EU representative: [OWNER: EU representative (GDPR Art. 27), if required]. UK representative: [OWNER: UK representative (UK GDPR Art. 27), if required].
When you produce a show, you decide whose voices and details go into it (guests, co-hosts, listeners on a private feed). For that personal data you are the controller and we process it for you as your processor, under our data processing addendum.
2. What we collect
- Account: name, email and avatar from sign-in (Clerk), plus your timezone, language and workspace settings, and the teammates you invite.
- Your work: shows, episodes, briefs, scripts, Show Bible, voice assignments, recorded takes, transcripts, generated audio, mixes, masters, cover art, media library files and RSS feeds.
- Your voice, only if you make one (see section 5).
- Guests: if you are a guest, the name and email the show used to invite you, the bio, links, photo and name pronunciation you enter, your recordings, and your acceptance of the release (typed name, date and release version).
- Connections and keys: your own provider API keys and your YouTube authorisation token, encrypted with AES-256-GCM. Only the last four characters of a key are ever shown back. Public API keys, private-feed links, guest links and AppSumo licence keys are stored only as hashes.
- Usage: credits spent, features used and API request counts, so we can bill fairly and enforce plan limits.
- Billing: plan, invoices and payment status from Stripe. Card details stay with Stripe. If you redeem an AppSumo lifetime deal: your licence key (hashed) and tier.
- Support: what you send us when you write in or report a problem.
- Referrals: if you arrive through a referral link, the referral code (in a cookie for 30 days) so the referrer can be credited.
- Analytics and diagnostics, only with your consent (see section 7).
- Technical logs: our servers record requests (IP address, user agent, time, path) to keep the service secure and working. [OWNER: confirm server log retention period].
3. Why we use it, and our legal bases
| Purpose | Legal basis (GDPR / UK GDPR) |
|---|---|
| Running your account and producing, storing and publishing your episodes | Contract (Art. 6(1)(b)) |
| Billing, credits, plan limits | Contract |
| Tax and accounting records | Legal obligation (Art. 6(1)(c)) |
| Security, abuse and fraud prevention, rate limits, server logs | Legitimate interests (Art. 6(1)(f)) in a safe service |
| Smart-link click counts and AI usage metrics (no content) | Legitimate interests in understanding and improving the service |
| Product analytics, Session Replay, error reports | Consent (Art. 6(1)(a)), given in the cookie banner |
| Your AI voice | Explicit consent (Art. 9(2)(a)), given by your spoken statement |
| Service emails you asked for (renewal reminders, production notices) | Contract |
We do not send marketing email without your consent and we do not make decisions with legal or similarly significant effects about you by automated means alone.
4. AI and your content
We do not use your content to train AI models. We send each provider only what the step you started needs: text to be voiced, audio to be transcribed, a prompt for cover art. Where a provider offers an opt-out from model training or a no-storage mode, we turn it on (for example Deepgram's model-improvement opt-out and OpenAI requests with storage off). Requests made with your own API keys follow your agreement with that provider.
AI voices are synthetic. Episodes made with them carry an AI marker in their audio metadata and feed. See the AI policy.
5. Your voice (biometric data)
If you choose to make an AI copy of your own voice (Settings → Your voice), you read a consent statement with three code words and three short passages, live. Deepgram transcribes the readings to check you read them; ElevenLabs creates the voice from them. Voice recordings and a voice model can count as biometric data under laws such as the GDPR (Art. 9), Illinois BIPA, Texas CUBI and Washington's biometric law. We treat them that way:
- Consent first. We create a voice only after your spoken consent statement and a written confirmation. The full terms are in the voice clone consent.
- One purpose. Your voice is used only to voice your own shows, only by you (not teammates). It is never sold, leased or traded, never used for anyone else and never used to train AI models. It is disclosed only to ElevenLabs and Deepgram, as our processors, to create and run it.
- Deletion. Deleting the voice, or your account, removes the ElevenLabs copy and your recordings straight away. Unfinished voices and their recordings are deleted after 7 days.
- Proof of consent. When a voice is deleted we keep a minimal consent record as proof that you agreed: a one-way hash of your account ID, of the IP address you consented from and of the code words, a fingerprint (SHA-256) of the consent recording, the version of the consent terms, and when you consented and when the voice was deleted. It holds no audio, no name and nothing that can recreate or identify your voice, and it is kept for 3 years after the voice is deleted, including after you delete your account, then deleted. [COUNSEL: confirm this period]
- Retention limit. We destroy voice data once its purpose is over, and in any case once the voice has not been used for 3 years. We tell you 30 days before, in CastGenX and by email; using the voice once keeps it.
6. YouTube
If you connect a YouTube channel, CastGenX uses the YouTube API Services. We ask Google for two permissions: to upload videos and to read your channel's basic details and the status of videos we uploaded. We use them only to publish the episodes you choose and to show whether YouTube has finished processing them. We store an encrypted refresh token and the IDs and status of uploaded videos; we do not read your other videos, comments or analytics.
You can disconnect in CastGenX at any time; we then revoke the token with Google and delete it. You can also remove access in your Google security settings. Google's use of your data is described in the Google Privacy Policy, and uploads are subject to the YouTube Terms of Service.
7. Analytics, Session Replay and error reports
Only if you choose “Allow all” in the cookie banner, we load Amplitude (product analytics and Session Replay) and Sentry (error reports). Amplitude records pages viewed, clicks and similar interactions, device and browser details and approximate location from your IP address. Session Replay records, for about 20 % of visits, a reconstruction of how our public website and legal pages were used, with all text on the page and everything you type masked; it never records the studio, settings, guest or invite links, or shared reports. Sentry records technical details of errors. “Essential only” keeps all of them off, and so does a Global Privacy Control signal from your browser. You can change your choice on the cookie policy page.
Separately, our server records metrics about AI text-model calls (which feature, model, token counts, cost, speed, errors and your account ID) in Amplitude in metadata-only mode. No prompt, script or reply text is included.
8. Listeners, feeds and smart links
- Public feeds: podcast apps fetch your feed and audio from our servers like any website, which sees the request's IP address and app.
- OP3 download counts: if a show turns on OP3, download requests for its episodes pass through op3.dev, an open-source measurement service, which sees the listener's IP address and app. We receive only totals by day, country and device.
- Private feeds: each listener link is stored as a hash with a label the show chose and when it was last used.
- Smart links: when someone follows a smart link we add one to that link's daily click count. We read the browser type only to leave out bots; we do not store IP addresses or identify who clicked.
- Feed update pings: when a public feed changes we send its URL (and nothing else) to Podcast Index and Podping. Private feeds are never announced.
9. Who processes it
Producing an episode means sending parts of it to the providers below. Each receives only what its step needs.
- Clerk: Sign-in and account identity.
- Deepgram: AI voices (text to speech) and transcription. Requests opt out of Deepgram's model-improvement program.
- NVIDIA Build: Script writing and editing suggestions (the default text model).
- OpenAI: Text generation and cover art, when chosen or when you add your own key. Requests are sent with storage turned off.
- ElevenLabs: Premium voices, AI music and AI sound effects, only if you use them.
- TypeSafe: AI quality and safety checks on script and metadata text (no audio is sent).
- Stripe: Payments, subscriptions and invoices.
- Resend: Account and production emails you've asked for.
- Sentry: Error reports, only if you choose “Allow all” in the cookie banner.
- Amplitude: Product analytics and Session Replay in the browser, only if you choose “Allow all” in the cookie banner. On our server it also records usage metrics for AI text-model calls (model, token counts, cost, speed and errors) in metadata-only mode, without any prompt, script or reply text.
- Google (YouTube): Uploading episodes to YouTube, only if you connect your YouTube channel. Uses the YouTube API Services.
- OP3 (Open Podcast Prefix Project): Open download counting for a show, only if you turn it on in the show's analytics settings.
- Podcast Index and Podping: Tells podcast apps that a public feed has a new episode. Private feeds are never announced.
- AppSumo: Lifetime-deal licence redemption.
The app, the database and your files run on our own server. The full list, with what data each provider receives, where, and a link to its data processing terms, is on the subprocessors page.
Otherwise we share personal data only: where you publish it (your feed, directories, YouTube); with your workspace teammates; when the law requires it or to protect rights and safety; or with a buyer or successor if CastGenX is sold or merged, under this policy. We do not sell personal data.
10. International transfers
Several providers process data in the United States. When personal data from the EEA, the UK or Switzerland goes to a country without an adequacy decision, we rely on the EU-US Data Privacy Framework (and its UK and Swiss extensions) where the provider is certified, and otherwise on the European Commission's Standard Contractual Clauses with the UK International Data Transfer Addendum, plus extra safeguards such as encryption in transit. You can ask us for a copy of the relevant safeguards. [OWNER: confirm the server location and that each provider DPA with SCCs is signed].
11. How long we keep it
| Data | How long |
|---|---|
| Account, shows, episodes and files | While your account is open. Deleted items wait 30 days in the Trash, then are purged. |
| Everything, on account deletion | Deleted at once, including your sign-in at Clerk, your voices and your keys. |
| Your voice and its recordings | Until you delete the voice or your account, or it goes unused for 3 years (notice 30 days before). Unfinished voices: 7 days. |
| Voice consent record (hashes and dates, no audio or name) | 3 years after the voice is deleted, also after account deletion (see section 5). |
| Guest details and recordings | Guest links expire after 14 days; the guest's details and takes stay with the episode until the show deletes them. |
| Private-feed links | Until revoked or the show is deleted. |
| Smart-link click counts | Until the link is deleted. |
| YouTube token | Until you disconnect YouTube or delete your account. |
| Payment and tax records | As long as tax law requires. [OWNER: confirm period, typically 6–10 years] |
| Amplitude analytics and replays (with consent only) | [OWNER: set in Amplitude and confirm; suggested 12 months for events, 30 days for replays] |
| Sentry error reports | [OWNER: confirm Sentry retention] |
| Server logs and backups | [OWNER: confirm periods] |
12. Your rights
Depending on where you live, you can ask to access, correct, delete or export your personal data, to restrict or object to how we use it, and to withdraw consent at any time (this does not affect what we did before). Most of this you can do yourself:
- Export: Settings → Profile → Export your data downloads everything we hold about you, with short-lived links to your media.
- Delete: Settings → Profile → Delete account permanently removes your shows, audio, feeds, voices, keys and sign-in.
- Correct: edit your profile and content at any time in the studio.
- Withdraw consent: change your cookie choice on the cookie policy page, or delete your voice in Settings → Your voice.
For anything else, email support@castgenx.com. We may ask you to confirm it is you, and we reply within one month (45 days under US state laws), telling you if we need longer. If you are a guest or a listener of a show, we may pass your request to the show, which controls that data.
If you are in the EEA or the UK you can complain to your data protection authority (in the UK, the ICO). US residents may use an authorised agent and may appeal a refusal by replying to our answer; if we still refuse, you can contact your state attorney general. We never treat you differently for exercising your rights.
13. Notice for California and other US states
In the last 12 months we collected the categories below, from you, your device, guests and teammates you invite, and our providers. We use them for the purposes in section 3 and disclose them only to the processors in section 9, for a business purpose.
| Category | Examples | Disclosed to |
|---|---|---|
| Identifiers | Name, email, account ID, IP address | Clerk, Stripe, Resend, Amplitude, Sentry |
| Customer records | Billing name and address | Stripe |
| Commercial information | Plan, purchases, credits | Stripe, AppSumo |
| Internet activity | Pages, clicks, session replays (with consent), API usage | Amplitude, Sentry |
| Audio and electronic information | Recordings, takes, generated audio | Deepgram, ElevenLabs |
| Biometric information | Your voice clone and its recordings (only if you make one) | ElevenLabs, Deepgram |
| Approximate location | Country or city from IP address | Amplitude |
| Professional information | Guest bios and links you enter | Published where the show publishes |
We do not sell personal information and we do not share it for cross-context behavioural advertising. We have not done so in the last 12 months and we have no actual knowledge of selling or sharing data of anyone under 16. We use sensitive personal information (your voice, account sign-in) only to provide the service you asked for. Retention is in section 11. See your privacy choices.
14. Children
CastGenX is not for children. You must be at least 16 to have an account, and the service is not directed at children under 13. If we learn we hold personal data of a child under 16 without valid consent, we delete it. If you believe a child has given us personal data, contact support@castgenx.com.
16. Changes
We update this policy when our processing changes and show the date at the top. For material changes we tell you by email or in the app before they take effect.
17. Contact
Questions or requests: support@castgenx.com, or write to [Company legal name — OWNER], [OWNER: registered postal address].