Privacy Policy
Effective date: November 4, 2026. The AudioStem.ai service is operated by J3X Solutions LLC ("we", "us"); AudioStem.ai is a product and brand of J3X Solutions LLC. This Privacy Policy explains what information we collect through AudioStem.ai, how we use it, who we share it with, and the choices you have.
The short version: we collect only what we need to run an audio service, we never sell your data, and we never use your audio to train AI models.
1. Information we collect
Account: your email address and name (if provided), plus authentication identifiers. If you sign in with Google, see the "Google Sign-In" section below for exactly what Google sends us.
Content: the audio you upload or link, the lyrics, descriptions, and settings you give for AI generation, the stems, songs, and other outputs we generate for you, and job metadata (model used, durations, timestamps, and your rights attestation). If you use the writing helper ("Write it for me (AI)", "Suggest a voice (AI)", "Write lyrics (AI)"), the few words or the video title you ask it to work from, and the style, language and length you chose, are the text it works from; we do not store that text or the suggestion it returns, only counters and token totals.
Usage and logs: standard technical records (IP address, browser and device information, request and error data) kept for security and operations.
Payments: handled entirely by Stripe, for subscriptions and for one-time purchases alike. We store your Stripe customer identifier, your subscription identifiers and plan status, and for each one-time purchase (such as a track download unlock) the Stripe checkout and payment identifiers, the track it applied to, when it was paid and when its period ends; card numbers never touch our systems.
2. Google Sign-In
What we receive: if you choose "Continue with Google", Google sends us your name, your email address, whether Google has verified that email, and your Google account identifier. Google’s standard profile scope also transmits a link to your profile image, but we do not store it. The image is discarded, not saved to your profile.
Why we use it: authentication and account management only. That means creating your account, signing you in, and linking your Google identity to an existing AudioStem.ai account with the same verified email. Nothing else.
Automatic account linking: we identify accounts by email address, which is the unique identifier for a user on the Service. If you sign in with Google using an email that already belongs to an AudioStem.ai account with that same verified email, we automatically link the two into one account instead of creating a duplicate, so you reach the same profile, files, and subscription whichever way you sign in. For your protection, if the existing account’s email is not verified we do not link automatically (you must sign in with your password first). By using both a password and Google on the same email address, you consent to this linking.
Sharing: Google-derived data is never sold. It is handled by Amazon Web Services, the infrastructure processor that hosts our user directory, as part of operating the Service; and if you make any purchase (a paid plan or a one-time purchase such as a track download unlock), your email address is sent to Stripe, our payment processor, to set up and manage your subscription or to record that purchase and send its receipt. No other third parties receive it — in particular, none of it is ever included in a request to the writing helper’s model provider. It is not used for advertising and is not used to train AI models.
Storage and retention: the data we keep (name, email, verification status, account identifier) is stored in our user directory (AWS Cognito) and your account profile, and is retained only for as long as your account exists.
Deletion: delete your account from the Account page for immediate, permanent erasure of your uploads, stems, profile, and Google-derived profile data. One narrow carve-out: we retain minimal internal audit and deletion-evidence records of account events (such as linking your Google identity, and the deletion itself), which can include your account identifiers (your email address and Google account identifier) kept solely for security, legal, and audit purposes and never used for anything else. You can also send a verified deletion request to privacy@audiostem.ai.
AudioStem.ai’s use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy (https://developers.google.com/terms/api-services-user-data-policy), including the Limited Use requirements.
Questions about Google data: privacy@audiostem.ai.
3. How we use it
To provide the Service (process your audio, deliver stems, enforce plan limits), to bill you, to answer support requests, to keep the platform secure (fraud and abuse prevention), to send service emails (verification, receipts, important account notices), and to meet legal obligations.
Marketing email (product news and tips) goes only to people who are subscribed. You can subscribe in the welcome window shown when you first sign in, in the one-time prompt in the app, or with the Email preferences toggle in Settings. In the welcome window the product-news box may already be ticked, depending on your device’s time zone; you can untick it before continuing. It always starts unticked if your device’s time zone is in the EU/EEA, the UK, Switzerland, Canada or South Korea, or cannot be determined, and it never starts ticked if you previously declined. We record when and how you subscribe or unsubscribe, including whether the box started ticked. Separately, if you have an account we may occasionally send you a one-time offer about your own account (such as a free month), even if you have not subscribed to product news; every such email has the same one-click unsubscribe link, and once you unsubscribe we will not send you another. Every marketing email has a one-click unsubscribe link, and you can opt out at any time in Settings.
Track-ready emails tell you that a separation or AI song you started has finished. They contain no marketing. They are on by default for new accounts (the first box in the welcome window, which you can untick) and can be turned off at any time in Settings (Email me when my track is ready). Service and billing emails (verification, receipts, renewal reminders, and account or security notices) are sent regardless, because they are part of running your account.
4. AI processing of your audio and text
Your audio is processed by machine-learning models running on our own infrastructure, solely to produce the outputs you requested. Your audio and the stems, songs and vocals we generate from it are not used to train or improve models (ours or anyone else’s) and are never shared with model vendors.
The writing helper is different, and it is optional: when you press "Write it for me (AI)", "Suggest a voice (AI)" or "Write lyrics (AI)" — or use a YouTube link as a song’s reference sound, which writes the song’s style from that video’s title automatically — the short text you asked it to work from (your words, or the title of a YouTube video you pasted — never the video or any audio) is sent to Anthropic’s Claude API, which returns the suggestion. Nothing about your account (no email, name or identifier) is included. Anthropic does not use this text to train its models; under its commercial terms it may retain it for a limited period to detect abuse. The suggestion goes into the box in front of you, and you decide whether to keep, edit or discard it. Every suggestion is written by an AI.
5. Your voice ("My voice")
If you create a voice profile ("My voice"), we collect: the voice sample you choose (10–30 seconds of your own singing or speech, recorded with your microphone or uploaded as a file); the recording of you reading the check phrase we show you; a transcript of that recording; two similarity measurements from the check (how closely the words matched the phrase, and how closely the reading matched your sample — kept for the check and for investigating misuse, and never shown to you or anyone else); the microphone settings your browser reported (echo cancellation, noise suppression, automatic gain, sample rate and channel count — never a device identifier); the consent you gave and when you gave it; and a record of each time the profile is used, so every file made with it can be traced to it.
We use this data to confirm that the voice is yours, to sing in your voice when you ask, to keep a record of your consent, and to investigate reports of misuse. Where the law treats it so, a voice profile may be a biometric identifier (for example under the Illinois Biometric Information Privacy Act). We collect it only with your written consent, given in the app before the check, and we never sell, lease, trade or otherwise profit from it.
Retention schedule. A voice profile that does not pass the check is deleted, with its recordings, 24 hours after the last attempt. A profile that passes is kept until you delete it or — a proposed rule, shown in the app — until it has gone 12 months without being used. When a profile is deleted, by you or by that schedule, its recordings are removed and we keep a consent receipt (the version of the consent wording you agreed to, the timestamps, the check’s outcome and a fingerprint of the sample — no audio) for a limited period, so that audio made with the profile can be traced to a valid consent if a dispute arises. Audio you already made with the profile is yours and stays with your tracks.
Your voice sample, your check recording and the audio we generate in your voice are never used to train AI models, ours or anyone else’s, and are never shared with model vendors.
Delete your voice profile at any time from the Account page, or delete your account for a full purge. A data export (section 8) includes your voice profile’s record and your own recordings.
6. Who we share it with
Only the processors that run the Service: Amazon Web Services (hosting and storage), Stripe (payments), Google (only if you choose Google sign-in), and Anthropic (only the text you ask the writing helper to work from, as described in section 4). We may also disclose information when the law requires it or to protect the Service and its users. We do not sell or rent personal information, and we run no advertising.
7. Retention and deletion
An account that holds a storage allowance granted to it (Terms §11) is kept, with its account record, until you delete it; it is not removed for inactivity.
Content is kept under your tier’s storage rules; some artifacts expire automatically. If your subscription ends (by cancellation or failed payment) and you do not reactivate within the 30-day notice window, we permanently delete your stored content (uploads, stems, and other outputs); your account itself remains as a free account unless you delete it. Full account deletion is always available in-app: when you delete your account we purge your uploads, stems, and profile (including Google-derived profile data) and cancel billing. We keep minimal deletion-evidence and audit records of account events (which can include account identifiers such as your email address and Google account identifier) so we can prove the erasure happened and meet legal obligations; they are used for nothing else. A YouTube link you use as a song reference is stored with the audio we fetched, the link, the video’s title and channel, and your rights confirmation, as part of your content, until you remove it or delete the last song that uses it. Backups and logs live on bounded, rolling schedules and age out automatically.
8. Your rights
Export: email privacy@audiostem.ai to request a copy of your data. We verify each request and prepare an archive (originals, stems, and metadata) that you can download from your Account page once it is ready.
Deletion: delete your account in-app for a full purge, or send a verified request to privacy@audiostem.ai.
For access, correction, or objection requests (including rights under GDPR or CCPA where they apply), contact privacy@audiostem.ai. We verify requests, respond within legally required timelines, and never penalize you for exercising your rights.
9. Cookies and local storage
We use only essential storage: the session tokens that keep you signed in and basic interface preferences (like dark mode). While we compare two versions of our homepage, your browser also remembers which version you were shown until you close the tab, and your account records it if you sign up from that visit. It holds no identifier. No advertising cookies, no third-party analytics trackers.
10. Children
The Service is not directed to children under 13 and we do not knowingly collect their data. If you believe a child has provided us data, contact privacy@audiostem.ai and we will delete it.
Where our AI features are used by people under 18, we keep safeguards in place: every AI-written suggestion is labelled as written by an AI; text sent to or returned from an AI model passes content filters, and requests that name real artists or try to steer the model are refused; usage is rate-limited and monitored; and no account data is ever sent to a model provider. Questions about these safeguards: privacy@audiostem.ai.
11. International transfers
The Service is hosted in the United States (AWS, US East), and the writing helper’s model provider (Anthropic) processes its requests in the US. If you use the Service from elsewhere, your data is transferred to and processed in the US as described in this policy.
12. Security
Data is encrypted in transit (TLS) and at rest; access follows least-privilege rules; media is served through short-lived signed URLs; administrative actions are logged. No service can promise perfect security. If you discover a vulnerability, please report it to security@audiostem.ai. We will notify affected users of a qualifying breach as required by law.
13. Changes and contact
If this policy changes materially we will post the update here, refresh the effective date, and notify you by email for significant changes. Questions or requests: privacy@audiostem.ai.