Legal

Privacy Policy

Last updated: 26 October 2026 · Effective: 26 October 2026

This Privacy Policy explains how Generative AI Teknoloji Anonim Şirketi ("AI Studio", "we", "us" or "our") collects, uses, discloses and protects personal information when you use the AI Studio app for iPhone and iPad and the related services we provide through it (together, the "Service"). It is written to meet the EU and UK General Data Protection Regulation ("GDPR"), the Turkish Personal Data Protection Law No. 6698 ("KVKK"), the California Consumer Privacy Act as amended by the California Privacy Rights Act ("CCPA"), other US state privacy and biometric laws, and Apple's App Store Review Guidelines.

In short: to create the images, videos, music and voiceovers you ask for, we send your prompts, the photos and videos you upload and your settings to the AI providers named in Section 6. Every request is checked for prohibited content before anything is generated, and image and video results are checked again before you see them. When you use a photo in a generation, template or tool, real people in it must be verified first: you with a selfie, others with their own selfie or a link they confirm. Face verification uses Amazon Rekognition and Sightengine and keeps a reference photo and a face template on your account until you delete that character. We remove location data from your photos and videos before they leave your phone. We keep your creations in your library, which only you can see in the app, until you delete them, and files you upload for a generation or template are deleted after 7 days (photos and videos you attach to a Director brief stay with that conversation, Section 10). We do not sell your personal information and we do not use your content to train AI models.

1. Scope and Who We Are

The data controller responsible for your personal information is Generative AI Teknoloji Anonim Şirketi, İdealtepe Mah. Dik Sok. No:13, Daire 2, Maltepe, İstanbul 34841, Türkiye, email hello@aistudio.works.

This Policy applies to personal information we process through the Service, including the consent page we host for people you ask to verify their face (Section 3.4). It does not apply to third-party apps, websites or services that you reach through the Service, such as Apple's App Store or the apps you share your creations to; their own privacy policies apply.

"Personal information" means information that identifies, relates to or can reasonably be linked to you or your device. It does not include information that has been de-identified or aggregated so that it can no longer reasonably be linked to you.

2. Changes to This Privacy Policy

We update this Policy when our practices or the law change, and we change the "Last updated" date at the top. If a change is material, we will tell you in advance in the app or by email before it takes effect. Where the law requires your consent to a change, we will ask for it. The version in force when we process your information applies to that processing.

This version describes face verification and characters (Section 3.4), the result checks (Section 4.2), AppsFlyer (Section 6.7), how location data is kept off uploads (Section 3.5), the in-app controls for withdrawing consent (Sections 7.1 and 7.3) and the control for deleting your history (Section 7.1). The face verification described in this version applies to you only after you agree to it in the app; the other changes take effect on the effective date above.

3. Personal Information We Collect

3.1 Information you provide to us

  1. Photos and videos. The photos and videos you choose in the system photo picker, or send to AI Studio from another app through the iOS share sheet, so that we can use them to create the content you request. The app only receives the items you select; it does not have access to the rest of your photo library.
  2. Prompts and other text. The prompts, lyrics, titles, captions, scripts and other text you enter, and the options you choose in templates and tools, such as the model, duration, aspect ratio and style.
  3. Generated content. The images, videos, music, voiceovers and other results the Service creates for you ("Outputs").
  4. Director conversations. The messages you exchange with the Director, our AI video assistant, the photos and videos you attach to them and the videos it produces for you.
  5. Selfies and characters. If you want a real person to appear in what you create, that person is verified with a selfie: you with the front camera on your device, or someone else with their own selfie on your device or on their own phone through a link you send them. We keep the selfie as a reference photo and create a face template from it, both on your account as a "character", together with the name you give it. You can also add an AI character from a photo of a person who does not exist. Section 3.4 explains this in detail.
  6. Preferences and onboarding answers. Settings you choose in the app, such as the gender you select so that we can suggest matching templates, your model fallback setting, your notification choices and your consent choices, and the answers you give during onboarding.
  7. Email address (optional). If you choose to add your email address in the app, for example to receive free credits once you confirm it and to get AI Studio updates. You can remove it at any time in Settings.
  8. Purchases. Which subscriptions and credit packs you buy, when, their price and currency, store transaction identifiers and your credit balance. We never receive your full payment card details: all purchases are processed by Apple.
  9. Your communications with us. Emails you send us, including support requests, feedback, content reports and appeals (manual review requests) against a refused request, and anything you attach. Emails you start from inside the app include your user ID, app version, device model and iOS version so that we can help you. Content reports sent from the app carry the reason you choose, your words, the ID of the content you report, and your app version, device model, iOS version and language.

3.2 Information collected automatically

  1. Account and device identifiers. When you first open the app, we create an anonymous account and a random app user ID; you do not need a name or password. Your app user ID is stored in your iCloud Keychain and iCloud key-value storage, so your account and credits stay with you when you reinstall the app or use another device signed in to the same Apple Account. We also process the Firebase user and installation identifiers, the identifier for vendor (IDFV) and identifiers set by the analytics, attribution and purchase services described in Section 6.
  2. Device and app information. Device model, operating system and version, app version and build, language, region and time zone.
  3. Usage information. The screens and features you use, the templates and models you open, the generations you start and their outcome (completed, failed or refused), credits spent and refunded, purchase and paywall events, and timestamps. Analytics events do not include the text of your prompts, except Director conversations, which are analyzed as described in Section 4.3. Versions of the iOS app released between September and early October 2026 included Amplitude Session Replay, which recorded a replay of the app's screens for about one in four sessions; text-entry fields such as prompts were masked, but photos and results shown on screen were not. Current versions do not include it, and on 6 October 2026 we turned recording off for the older versions as well.
  4. IP address and approximate location. Our servers and the service providers in Section 6 receive your IP address when your device connects to them, and may derive an approximate location from it, such as your country or city. We do not collect your precise location (see Section 3.5).
  5. Advertising identifier. Only if you allow tracking when iOS asks, we and our advertising measurement partners can read your device's advertising identifier (IDFA). You can change this choice at any time in iOS Settings › Privacy & Security › Tracking.
  6. Notification tokens. If you allow notifications, a push token that lets us notify you when your content is ready, and, while a generation is running, a Live Activity token that updates its progress on your Lock Screen and in the Dynamic Island. These notifications and Live Activities may show the first words of your prompt.
  7. Device integrity signals. Apple DeviceCheck signals, through Firebase App Check, that help us confirm requests come from a genuine copy of our app. These signals do not identify you.
  8. Crash reports. We do not use a third-party crash reporting tool. If you choose to share analytics with app developers in iOS Settings, Apple provides us with crash and performance reports that do not identify you.

3.3 Information we receive from other sources

  • Apple: the status of your App Store subscriptions and purchases, including renewals, cancellations and refunds; and, through Apple's AdServices framework, whether you installed the app after tapping an Apple Search Ads ad and, if so, the campaign and keyword.
  • Advertising and attribution partners: information about the ad campaigns and links that led to installs and purchases, which is usually aggregated.
  • People you ask to verify their face: their selfie and consent record, as described in Section 3.4.

3.4 Faces and biometric information

Photos you upload often show people's faces. To make sure that real people appear in your creations only with their consent, the Service verifies faces. This involves biometric information, so please read this section carefully.

When a photo is checked. When you start a generation, template or tool that uses a photo, the app sends the photo to our servers and Amazon Rekognition finds the faces in it. Photos attached to a Director brief are checked for prohibited content (Section 4.2) but not for faces. Each face that is large enough to be a subject of the photo is compared with the face templates of your characters. A face that matches one of your characters passes. A face that matches a public figure in Amazon's catalogue is refused, and no selfie can clear it. If any other face remains, Sightengine estimates whether the photo is AI-generated; if it is, those faces pass. A real face that is not one of your characters blocks the generation until that person is verified or a member of our team approves it on appeal. If our team approves your appeal, the face in that photo is added to your characters as a verified person, with a reference image and a face template, and you can delete it like any other character; we approve only when the photo shows you or a person whose consent you have shown us. If the AI-generated check cannot be completed and no public figure was found, the photo is allowed through so that an outage does not block you. A fingerprint (hash) of a photo that passed, stored with your user ID, lets the same photo pass again for one hour without a new check; the record is deleted when you delete a character or your account. The app calls this check "Verification" and shows you each face it found and what is needed for it.

Verifying yourself. You take a selfie with the front camera inside the app. The selfie is sent to Amazon Rekognition, which checks that it shows one clear face, compares it with the face in your photo, and creates a face template of it. We keep the selfie as the character's reference photo, a crop of the matching face from your photo, and the face templates (numerical representations of a face that Amazon Rekognition uses for matching; they are not images). You can add more photos of the person to improve matching; a character holds up to ten face templates in total.

Verifying someone else. Either the person takes their selfie on your device, or you send them a link. The link opens a page we host, shows them the face from your photo and asks "Is this you?"; they take a selfie with their own camera, it is compared with the face in your photo in the same way, and, if it matches, the character is created on your account. The link works once and expires after 7 days; the copy of your photo made for it is deleted after 7 days. The page tells them that their selfie is compared with the face in your photo and saved to your account as the face they allow you to use, and that we keep a record of their consent. For that record we store the time, a hashed version of their IP address, their browser type and a random identifier for their browser session, and we keep it for 5 years as proof that consent was given, even after the character is deleted. We do not create an account for them. Besides the selfie, which becomes the character's reference photo on your account (you can see it in Characters), we store only that record and the name you gave the character. Section 7.7 explains their rights.

AI characters. You can add a character from a photo of a person who does not exist. The photo is sent to Amazon Rekognition and Sightengine to confirm that it is not a public figure and that it is AI-generated; a photo of a real person is refused. The photo is kept as the character's reference photo with its face template.

What we keep and for how long. For each character: the name you gave it, the reference photo, the face crops taken from your photos, and the face templates, which live in a collection that belongs to your account alone at Amazon Rekognition. They are kept until you delete the character (Characters › delete, or Settings › Your Data › Remove All Characters) or your account, and in any case no later than one year after your last use of the Service. Deleting a character deletes its reference photo, crops and face templates. Photos uploaded only for a check are deleted after 1 day. If a check refuses a photo or selfie, a copy of it is kept for 30 days with the refusal record so that you can appeal, whether or not you do (Section 4.2).

Who processes it. Amazon Web Services, Inc. (Amazon Rekognition, US East region) performs face detection, comparison, template creation, template matching and public-figure detection. Sightengine SAS (France) estimates whether a photo shows an AI-generated person. We send them only the image being checked (and, to Sightengine, a link or the bytes of that image). They return the results to us. They may keep the images only as long as their terms with us allow and, as Section 6.10 says, may use them only to provide this service to us. Your reference photos and face crops are stored in our Google Cloud storage. We do not share your face templates with anyone else, we do not use them for any purpose other than verifying consent, matching characters and recognizing public figures so that they are not used, and we never sell, lease or trade them.

Your choice. Creating a character is optional. The face check itself runs on every photo you use that shows a face; if you do not want that, use templates and models without photos of people. Before you take a selfie, the app tells you that it is sent to Amazon Rekognition and kept with the character, and asks you to continue. By continuing you consent to this processing. You can delete characters at any time.

Photo Restore. The Photo Restore tool uses Apple's on-device Vision framework to find faces in your photo, to check that the photo is suitable for restoration, and to compare the shape of the face before and after restoration so that we can warn you if a face may have changed. That processing happens entirely on your device; its results are not stored and are not sent to anyone.

3.5 Location data

We do not collect your precise location and the app never asks for it. Photos and videos can carry location data (GPS tags) inside the file. Before a photo or video leaves your device, the app removes that location data (and, for videos, other identifying file metadata); the picture or video itself is unchanged. If a file is in a format the app cannot process, it is uploaded as it is.

3.6 What we do not collect

AI Studio uses your camera only to take a selfie for face verification (Section 3.4), and only when you choose to. It does not use your microphone, and it does not access your contacts, calendars, precise location, health data or the parts of your photo library you have not selected. When you turn on Closed Captions + SDH in iOS Accessibility settings, the app recognizes speech in videos to show captions; that recognition runs on your device and nothing is sent to us. The app's technical files declare some other permissions only because a component built into the app requires the declaration; the app never asks for them and never uses them.

4. How We Use Your Information

4.1 To provide the Service

  • to create your account and keep it available on your devices;
  • to send your request to the AI provider that runs the model you chose and return the result to you, including retrying a failed request on a backup model when model fallback is on;
  • to verify faces and keep your characters, as described in Section 3.4;
  • to show your creations in your private library and let you save, share and delete them;
  • to run the Director;
  • to calculate prices, deduct, grant and refund credits, process purchases, and restore subscriptions;
  • to send you notifications that your content is ready and show its progress in Live Activities.

4.2 To keep the Service safe

  • Before anything is generated, your prompt is screened by our own software for words that describe prohibited content, and your prompt and every photo and video attached to the request are checked by an automated classifier run by EachLabs for prohibited content: sexual or nude content, graphic violence and gore, hateful content, anything that sexualizes minors, and kissing, hugging or other intimate contact between people. A request to take the clothes off a person in a photo, or to put them in swimwear, underwear or lingerie, is refused as well, and so is a request to show anyone in swimwear, underwear or lingerie. For Virtual Try-On, each item photo you add is also checked by Google's Gemini model, and swimwear, underwear and lingerie are refused. The Director's own prompts for its images and clips are checked the same way before anything is generated. Requests that are flagged are refused (the app says "declined" or "blocked"), and you are not charged for them. If a check cannot answer, the request is refused and you are not charged. The AI providers also apply their own safety filters and may decline a request; if one declines it on safety grounds, the request ends and your credits are returned. If a provider fails for another reason, we try the next provider that runs the same model (and a backup model if model fallback is on), and if none succeeds your credits are returned.
  • After a result is generated, every image and video (including the Director's) is checked by the same classifier before it is shown to you. A result that breaks our rules is withheld: you do not see it, your credits are returned, and the file is kept in our storage at an unlisted link that is not shown in your library, together with a record our team uses to prevent abuse. If the classifier cannot answer for a finished result, the result is delivered, because the request already passed the checks before generation and the provider's own filter.
  • When a request or result is refused, we record the prompt, links to the attached files, the reason, the classifier's label, the name of any public figure recognized in a photo, and your user ID, so that we can handle appeals and prevent repeated abuse.
  • Appeals (manual review). When a request is refused and you can appeal, the app shows a "Request manual review" button that opens an email to us with a reference code, and we copy the photos from that request so that a reviewer can see them; the copies are deleted after 30 days whether or not you appeal. If you send the email, a member of our team reads your request and the related record and replies to you, usually within 48 hours. Every appeal is reviewed by a person, not automatically.
  • Reports. When you report content from the app, we store the report with a reference code and email it to our team, who review it within 24 hours.
  • We investigate reports, suspend or block accounts that break our Terms of Service (a blocked account can no longer generate), prevent fraud and abuse of free credits, and report child sexual abuse material to the authorities where the law requires.

4.3 To understand and improve the Service

To measure which features, templates and models are used, find and fix errors, and plan improvements. For the Director, this includes analyzing the text of your conversations and the Director's replies, keyed to your user ID, after an automatic filter removes personal details such as email addresses and phone numbers. The filter can miss some details, so please do not put personal details in Director chats.

4.4 To communicate with you

  • to answer your emails, support requests, reports and appeals;
  • to send service messages about your account, purchases and generations;
  • to send promotional notifications about offers and credits, only if you allowed notifications and did not turn off Marketing Notifications in Settings. Allowing notifications turns Marketing Notifications on; you can turn them off at any time in Settings › Account Settings;
  • to email you a link to confirm your email address, and AI Studio updates, only if you gave us your email address for that purpose. You can stop these emails at any time by removing your email address in Settings.

4.5 To measure our advertising

To learn which of our ads and links lead to installs, subscriptions and purchases, and to report conversions to the ad platforms we use. We use your advertising identifier for this only if you allowed tracking. We do not show third-party ads in AI Studio.

4.6 To meet legal obligations and protect rights

To keep tax and accounting records, respond to lawful requests from authorities, enforce our Terms, and establish, exercise or defend legal claims.

4.7 AI model training

We do not use your photos, videos, selfies, prompts, Director conversations or Outputs to train AI models. We send your content to AI providers only to create the result you requested or to run the checks described in this Policy.

4.8 Legal bases for processing (EEA, UK, Switzerland and Türkiye)

PurposeMain information usedLegal basis
Creating your content and running your account, library, credits and purchasesIdentifiers, uploads, prompts, Outputs, Director conversations, purchases, device informationPerformance of our contract with you; for sending your content to AI providers, also your consent on the consent page
Checking the faces in a photo you use (finding faces, comparing them with your characters and with Amazon's public-figure catalogue, estimating whether the photo is AI-generated)The photo and the faces in it, including people who are not youYour explicit consent on the consent page, which names this check (Article 9(2)(a) GDPR; KVKK Article 6). Nothing from the check is kept about people in your photo unless they are verified; the check exists to protect their rights
Face verification and charactersSelfies, reference photos, face templates, consent recordsYour explicit consent before you take a selfie (Article 9(2)(a) GDPR; KVKK Article 6); for the person who confirms through a link, their explicit consent on that page
Content checks before and after generation, refusal records, appeals, reports and enforcementPrompts, uploads, Outputs, user ID, refusal records, your messagesOur legitimate interest in keeping the Service lawful and safe; legal obligations
Analytics and product improvementUsage information, identifiers, device information, redacted Director conversationsOur legitimate interest in understanding and improving the Service
Advertising measurement using your advertising identifierIDFA, purchase and install eventsYour consent, given in the iOS tracking prompt
Other advertising measurementInstall and purchase events, IP address, device informationOur legitimate interest in measuring our marketing
Promotional notifications and emailsPush token, email addressYour consent, given when you allow notifications or add your email address, which you can withdraw in Settings
Support and communicationsYour messages and contact detailsPerformance of our contract; our legitimate interest in answering you
Tax, accounting and responses to authoritiesPurchase records, information requested by lawLegal obligations

Where we rely on legitimate interests, we have balanced them against your rights. You can object to this processing as described in Section 7.

AI Studio creates content with AI models run by third-party providers. Before you create anything, a consent page explains that your prompts, the photos and videos you upload and your settings are sent to the AI providers that run the models you choose, that faces in your photos are verified by Amazon Rekognition and Sightengine, that Director conversations are analyzed after an automatic filter removes personal details, names some of the providers, and asks for your agreement. The full list of providers is in Section 6. Before you take a selfie, the app tells you where the selfie goes and what is kept (Section 3.4). You can withdraw your consent at any time in Settings › Your Data, as described in Section 7.3. Withdrawing consent does not affect processing that happened before.

6. How We Disclose Your Information

We disclose personal information only to the recipients below and only for the purposes in this Policy. Our service providers process it on our behalf under contracts that limit their use of it to providing their services to us.

6.1 AI generation providers

When you start a generation, your prompt, the files you attached and your settings are sent to the provider that runs the model you chose. If that provider fails or declines, the request may be sent to another provider that runs the same model. If model fallback is on, it may also be finished by a backup model (Wan 3.0 for video, Imagine 2.0 for images); you can switch model fallback off in Settings › Account Settings › Model Fallback.

ProviderWhat it does for usWhat it receives
EachLabsRuns image and video models and all templates; runs our content checks before and after generation; a fallback for upscalingPrompts, photos, videos, settings, generated results for checking
KIE.aiRuns image, video and music models, the lyrics writer, music style suggestions and the image and video upscalersPrompts, lyrics, photos, videos, settings
Pika (Mellis, Inc.)Runs image and video models, including the Seedance, MiniMax, Wan, FLUX and Nano Banana modelsPrompts, photos, videos, settings
fal.ai (Features & Labels, Inc.)Runs image and video models (including Studio 2.0 images, Kling motion control, Grok Imagine and MiniMax H3 Max), Subtitle Studio and the Director's transcriptionPrompts, photos, videos, audio, settings
Google (Gemini API)Runs Google's image and video models (Nano Banana, Gemini Omni, Veo) and the Director's planning, images and clips; checks the item photos added to Virtual Try-OnPrompts, photos, videos, Director messages and attachments, Virtual Try-On item photos
ElevenLabsCreates voiceovers for Text to Speech and the DirectorThe text to be spoken and the voice chosen

Through these platforms, your request may be processed by the developer of the model you chose, including OpenAI, ByteDance, Google, Alibaba, Kuaishou, MiniMax, Black Forest Labs, xAI, Meta, Microsoft, Suno, Topaz Labs and VEED.

6.2 Content safety

EachLabs's content safety classifier checks your prompts, photos and videos before generation and your image and video results after generation. Google's Gemini model checks the item photos you add to Virtual Try-On. Many of the AI providers above also apply their own safety filters and may refuse a request.

6.3 Face verification

Amazon Web Services, Inc. (Amazon Rekognition) and Sightengine SAS process the photos and selfies described in Section 3.4. Amazon Rekognition holds the face templates of your characters in a collection that belongs to your account alone.

6.4 Infrastructure and hosting

Google Cloud and Firebase (Google LLC) host our servers, database, file storage, anonymous sign-in, App Check and push notifications. Apple delivers push notifications and Live Activity updates and provides DeviceCheck.

6.5 Analytics

  • Amplitude (EU data center) receives usage events (including the category of a refused request, but never the prompt), device information, your app user ID and app settings such as your selected gender and subscription status, the text of Director conversations and replies after an automatic filter removes personal details, and the session replays from the older app versions described in Section 3.2(c).
  • Google Analytics for Firebase receives standard app events, such as first open, sessions and in-app purchases, device information and app identifiers, and your advertising identifier only if you allowed tracking.

6.6 Purchases and subscriptions

  • Apple processes all purchases in the app.
  • Adapty (Adapty Tech Inc.) manages subscription status, purchase history, paywalls and purchase attribution for us, and receives your app user ID, device information, IP address, the identifiers our analytics and attribution partners assign to your device and, if you allowed tracking, your advertising identifier. Adapty also matches installs to the links in our ads.
  • Supabase (Supabase, Inc.) hosts our copy of purchase events for reporting.

6.7 Advertising measurement

  • AppsFlyer (AppsFlyer Ltd.) measures which of our ads and links lead to installs and purchases. The app starts AppsFlyer only after you have answered the iOS tracking prompt. AppsFlyer receives your app user ID, device information, IP address, the identifier for vendor, app events such as first open, onboarding completed, first generation and your first purchase with its price, and, only if you allowed tracking, your advertising identifier. It may also receive subscription renewals and refunds from Adapty, and it collects Apple's SKAdNetwork reports, which do not identify you. AppsFlyer may forward purchase events to the ad platforms we use.
  • Meta (Meta Platforms, Inc.), through the Facebook SDK in the app, receives app installs, app opens and your first generation, with device information, IP address, an anonymous identifier and, only if you allowed tracking, your advertising identifier. The SDK is started only after you have answered the iOS tracking prompt. If you did not allow tracking, the events are marked accordingly and Meta limits how it uses them.
  • Apple provides campaign attribution through Apple Search Ads and SKAdNetwork.

Under some US state laws, disclosing information to ad platforms for ad measurement can count as "sharing" for cross-context behavioral advertising. See Section 11 for how to opt out.

6.8 Email and support

Google Workspace hosts our mailbox, hello@aistudio.works. Resend (Resend, Inc.) sends our emails, such as the link to confirm your email address, acknowledgements of appeals and content reports, account messages and any AI Studio updates you asked for. Our system reads incoming emails with an appeal subject automatically to match your reference code, store your message with the appeal and send you an acknowledgement; a person then reviews the appeal.

6.9 Legal, safety and business transfers

  • We disclose information when the law requires it, to respond to valid requests from courts and authorities, to report child sexual abuse material to the National Center for Missing & Exploited Children (NCMEC) or the competent authority, and to protect the rights, property and safety of our users, the public and AI Studio.
  • If we are involved in a merger, acquisition, financing or sale of assets, personal information may be transferred to the other party, which must continue to protect it as this Policy describes.
  • We may disclose information with your consent or at your direction, for example when you share an Output to another app.

6.10 The same protection

Every third party that receives your personal information from us, including the AI, content safety, face verification, analytics, advertising measurement, payment and infrastructure providers named above, receives only the information it needs for its task, may use it only to provide its service to us, and must protect it at least as well as this Policy and Apple's App Store Review Guidelines require.

6.11 No sale of personal information

We do not sell your personal information, and we do not disclose your photos, videos, selfies, face templates, prompts or Outputs to anyone for advertising.

7. Your Privacy Choices and Rights

7.1 In the app

  • Stop sharing your content with AI providers with the switch in Settings › Your Data › Share with AI Providers. While it is off, nothing new is sent to the AI providers, and anything you try to create asks for your consent first; taking a selfie for a character has its own notice and consent. The switch applies to the device you set it on.
  • Delete a single creation from your library, or delete your whole history in Settings › Your Data › Delete All Account History: your library, your Director projects with their conversations and videos, and the photos and videos you uploaded for them. Work that is still being made, or that finished in the last 10 minutes, is kept for now, and you can delete it afterwards. Your account, credits, subscription and characters are not affected. Some copies stay for the periods in Section 10: the records we need to prevent abuse, including withheld results and, until they are deleted 7 days after upload, the photos and videos a refused request used, the redacted Director conversations in our analytics (Section 4.3), and copies AI providers keep under their own terms. Deleting a single Director video from your library removes it from your library, but its video file and conversation stay until you use Delete All Account History or delete your account. Deleting a Director project that was never produced (the bin icon in its conversation) deletes the conversation and the photos and videos attached to it within minutes. If your version of the app does not offer Delete All Account History, update the app or ask us by email (Section 7.5).
  • Delete a character in Characters, or all of them in Settings › Your Data › Remove All Characters. This deletes their reference photos and face templates.
  • Turn off promotional notifications in Settings › Account Settings › Marketing Notifications.
  • Turn off model fallback in Settings › Account Settings › Model Fallback.
  • Remove your email address in Settings.
  • Stop AI Studio update emails by removing your email address in Settings.

7.2 On your device

  • Allow or deny tracking, and change your choice later, in iOS Settings › Privacy & Security › Tracking.
  • Turn notifications on or off in iOS Settings › Notifications › AI Studio.
  • Allow or deny camera access in iOS Settings › Privacy & Security › Camera. Without it you cannot take a selfie to verify a face, but everything else works.
  • The app works with the system photo picker, which needs no access to your library. Saving a result asks only for permission to add to your library.

7.3 Withdrawing consent

You can withdraw any consent you have given at any time. To stop sharing your content with AI providers, turn off Settings › Your Data › Share with AI Providers. To withdraw your consent to face verification, delete the character in Characters or use Settings › Your Data › Remove All Characters; a person who confirmed their face through a link can ask us to delete it as described in Section 7.7; we keep the record that the person consented, as proof, as described in Section 10. To withdraw consent to tracking, promotional notifications or emails, use the controls in Sections 7.1 and 7.2. Withdrawal does not affect processing that happened before it.

7.4 Your rights

Depending on where you live, you may have the right to:

  • access the personal information we hold about you and receive a copy of it;
  • receive your information in a portable format and have it sent to another company;
  • correct inaccurate information;
  • delete your information;
  • restrict or object to our processing, including processing based on legitimate interests;
  • opt out of the "sale" or "sharing" of personal information and of targeted advertising;
  • withdraw consent; and
  • complain to a data protection authority.

7.5 How to make a request

Email hello@aistudio.works or use Settings › Need Help › Contact Us in the app, and include your user ID, shown in Settings › Account Settings › User ID. Because accounts are anonymous, we verify requests by confirming that you control the account, for example by asking you to send the request from the app or to confirm details only the account holder would know. You may use an authorized agent where the law allows; we may ask the agent for proof of authority and ask you to confirm your identity. We respond within the time the law requires, for example within one month under the GDPR, 30 days under the KVKK and 45 days under the CCPA. If we decline your request, we will explain why, and where the law provides it you may appeal by replying to our decision; we will answer your appeal within the time the law requires.

7.6 Deleting your account

You can ask us to delete your account at any time by emailing hello@aistudio.works with your user ID, or by using Settings › Need Help › Contact Us. Within 30 days we delete your account and the personal information linked to it, including your creations, uploads, Director conversations, characters with their reference photos and face templates, and email address, except information we must keep by law, such as purchase records, to defend legal claims, or to prevent abuse: records of refused requests and withheld results, the record that a person consented to appear (Section 10), and the hash showing that an email address already received the one-time credits. We also ask Amplitude, AppsFlyer and Adapty to delete the data linked to your app user ID. Deleting your account does not cancel an App Store subscription; cancel it in your Apple Account settings first. Unused credits are lost when the account is deleted.

7.7 If you are not a user

If someone asked you to confirm your face through a link, or used a photo of you, you can ask us at any time what we hold about you and receive a copy of it, ask us to delete your face template, the reference photo and the character, or to stop a specific use, by emailing hello@aistudio.works. The person who created the character can see your reference photo in their Characters list until it is deleted. Tell us the name of the person who sent the link if you know it; we may ask you questions to confirm that the request is genuine, and we will not ask you for a new selfie. We answer within one month (45 days for US residents). We keep the record that you consented (time, hashed IP address, browser type) for 5 years as proof, unless you ask us to delete it and no legal claim requires it. You also have the rights in Sections 7.4 and 12, including the right to complain to a data protection authority.

8. Security of Your Information

We protect personal information with technical and organizational measures appropriate to the risk, including encryption in transit, access controls that limit production access to staff who need it, storage rules that let only your own signed-in device read your characters and face-check photos, unguessable file links, and Firebase App Check to help verify that requests come from genuine copies of our app. Face templates are held in a collection that belongs to your account alone. Your results and Director videos are stored at unguessable links so that the app can play them and you can share them; anyone who has such a link can open the file, so share links carefully. Deleting a result deletes its file; a Director video's file is deleted with its project (Section 7.1). No method of transmission or storage is completely secure, so we cannot guarantee absolute security. If a personal data breach affects you, we will notify you and the competent authorities as the law requires.

9. International Data Transfers

We are based in Türkiye. Our servers are hosted by Google Cloud in the United States, our product analytics are hosted by Amplitude in the European Union, face templates are held by Amazon Web Services in the United States, Sightengine processes images in France, and our AI providers, model developers, attribution and other service providers process information in the United States, the European Union, Israel and, for some models, the countries where the model developers named in Section 6.1 operate, which may include China and Singapore. These countries may not offer the same level of protection as your own. Where the GDPR, the UK GDPR, the KVKK or similar laws require it, we protect transfers with appropriate safeguards, such as the European Commission's Standard Contractual Clauses, the UK International Data Transfer Addendum and the standard contracts required by the KVKK. You can ask us for more information about these safeguards.

10. Retention of Personal Information

We keep personal information only as long as we need it for the purposes in this Policy, then delete or anonymize it.

InformationHow long we keep it
Photos, videos and audio you upload for a generation or template (Director attachments: see below)Deleted automatically 7 days after upload
Photos uploaded only for a face checkDeleted automatically after 1 day
Fingerprint (hash) of a photo that passed a face checkUsed for 1 hour; the record is deleted when you delete a character or your account
Characters: reference photos, face crops and face templatesUntil you delete the character or your account, and no later than one year after your last use of the Service
Face consent links and the copy of your photo and of the face made for themLinks work once and expire after 7 days; the copies are deleted automatically after 7 days
Consent records for a person who confirmed through a link (time, hashed IP address, browser type, browser-session identifier, the name you gave the character, link reference)5 years from the date of consent, as proof that consent was given, even if the character is deleted earlier; earlier on request if no legal claim requires it
Your creations (Outputs) and the prompts and settings stored with themUntil you delete them or your account
Director conversations, attachments and videosUntil you delete them (Settings › Your Data › Delete All Account History) or your account
Copies of videos made only for the content checkDeleted automatically after 1 day
Records of refused requests and withheld results (prompt, links to the files, reason, the classifier's label, face-check results including the name of a recognized public figure, the withheld results and the Director frames that were flagged, user ID)As long as needed to prevent abuse, handle appeals and meet legal obligations. The photos and videos uploaded for a refused request are still deleted 7 days after upload, and the copies made for a review after 30 days
Photos and selfies copied when a request or face check is refused, for a possible appealDeleted automatically after 30 days
Server logs (technical records of requests, including your user ID and face-check results)30 days
Content reportsAs long as needed to handle the report and any follow-up
Email confirmation linksValid for 7 days
Record that an email address has already received the one-time confirmation credits (a hash of the address, not the address)Kept permanently to prevent repeated claims
Account information, preferences, credits and email addressWhile your account exists
Purchase and transaction recordsAs long as tax and accounting laws require, generally up to 10 years
Analytics and attribution dataFor the retention periods configured in Amplitude, Google Analytics for Firebase and AppsFlyer
Support emails and appealsAs long as needed to handle your request and any follow-up

Deleted files may remain in encrypted backups or provider systems for a short period before they are overwritten. Our AI providers process your content to create your result; they may keep it for a limited time under their own terms, for example to detect abuse.

11. Supplemental Notice for United States Residents

This section applies to residents of California and other US states with comprehensive privacy laws, including Colorado, Connecticut, Delaware, Florida, Indiana, Iowa, Kentucky, Maryland, Minnesota, Montana, Nebraska, New Hampshire, New Jersey, Oregon, Rhode Island, Tennessee, Texas, Utah and Virginia, and supplements the rest of this Policy.

11.1 Categories of personal information

In the past 12 months we collected the categories below from you, your device and the sources in Section 3.3, for the purposes in Section 4, and disclosed them for a business purpose to the categories of recipients shown.

CategoryExamplesDisclosed to
IdentifiersApp user ID, Firebase and device identifiers, IP address, email address (if provided), advertising identifier (if allowed)Infrastructure, analytics, purchase, attribution and advertising measurement providers
Commercial informationSubscriptions and credit packs bought, credit balancePurchase providers, analytics, attribution and advertising measurement
Internet or other electronic network activityApp usage, features used, generation outcomes, device informationAnalytics and infrastructure providers, attribution and advertising measurement
Audio, electronic, visual or similar informationPhotos and videos you upload, OutputsAI generation and content safety providers, face verification providers, infrastructure providers
Biometric informationSelfies, reference photos and face templates of your charactersFace verification providers (Amazon Rekognition, Sightengine), infrastructure providers
Approximate geolocationCountry or city derived from IP addressInfrastructure, analytics, attribution and advertising measurement providers
Characteristics of protected classificationsThe gender you select for template suggestions (optional)Infrastructure and analytics providers
Other content you providePrompts, Director conversations, messages to usAI generation and content safety providers, infrastructure, analytics (Director, redacted), email providers

Of these, the biometric information in Section 3.4 is sensitive personal information under the CCPA. We use it only to verify that the people in your photos agreed to appear, to match your characters and to keep the Service safe, which are uses necessary to provide the Service you request; we do not use it to infer characteristics about you and we do not disclose it for any other purpose. We do not collect other sensitive personal information such as precise geolocation, government identifiers or account log-in credentials. We keep each category for the periods in Section 10.

11.2 Sale and sharing

We do not sell personal information for money. Disclosing identifiers, commercial information and network activity to Meta, AppsFlyer and similar ad platforms for ad measurement may be considered "selling" or "sharing" for cross-context behavioral advertising, or "targeted advertising", under some state laws. To opt out, deny tracking in iOS Settings › Privacy & Security › Tracking (which we treat as an opt-out), or email us with the subject "Do Not Sell or Share My Personal Information". We never sell or share biometric information, and we do not knowingly sell or share the personal information of consumers under 16.

11.3 Your rights

Subject to the law of your state, you have the right to know what personal information we collect, use and disclose; to access it and receive a portable copy; to correct it; to delete it; to opt out of its sale, sharing and use for targeted advertising; and, where applicable, to obtain a list of the specific third parties to whom we disclosed it. You can make a request as described in Section 7.5. We will not discriminate against you for exercising your rights.

11.4 Biometric privacy notice

This notice is given under the Illinois Biometric Information Privacy Act, the Texas Capture or Use of Biometric Identifier Act, Washington's biometric identifiers law and similar laws.

  • What we collect. When you verify a face, Amazon Rekognition creates a face template (a numerical representation of the face used for matching, not an image) from the selfie and from the matching face in your photo, or, if our team approves an appeal, from the face in the photo you appealed. We also keep the selfie as the character's reference photo, together with the face crop.
  • Purpose. Only to confirm that a real person in your photo agreed to appear in what you create, to recognize that person in your later photos, to keep people who did not agree out of your creations, and to recognize whether a face is a public figure in Amazon's catalogue so that it is not used; when a public figure is recognized, the catalogue name is kept in the refusal record. We do not use biometric information to identify you or anyone else for any other reason.
  • Consent. We collect biometric information only after telling you, in the app, what is collected and why, and after you choose to continue. A person who confirms through a link gives their consent on that page.
  • Retention and destruction. We permanently destroy a character's face templates and reference images when you delete the character or your account, and in any case no later than one year after your last use of the Service. Deleting the character also deletes the templates from Amazon Rekognition.
  • No sale, no profit. We never sell, lease, trade or otherwise profit from biometric information.
  • Disclosure. We disclose biometric information only to Amazon Web Services and Sightengine as our processors for this feature and to Google Cloud, which stores the reference photos and face crops, when you consent or direct it, or when the law requires it.
  • Security. We protect it using the reasonable standard of care in our industry and at least as protectively as we protect other confidential information, as described in Section 8.

The on-device face detection in Photo Restore described in Section 3.4 never leaves your device and is not used to identify anyone.

11.5 Other notices

  • Nevada: we do not sell covered information as defined by Nevada law. You may still email us to submit an opt-out request.
  • California "Shine the Light": we do not disclose personal information to third parties for their own direct marketing purposes.
  • Do Not Track: the app does not respond to Do Not Track signals. It uses your advertising identifier for ad measurement only if you allow tracking in the iOS prompt (Section 4.5).
  • Financial incentives: we may offer free credits when you add and confirm your email address or send us feedback. The terms are shown in the app before you accept. We currently offer 20 free credits when you confirm your email address. We estimate the value of your email address to us at about the price of 20 credits in our smallest credit pack, because that is what we spend to be able to tell you about new features and offers; the estimate is calculated from the price of the credits given. You can withdraw at any time by removing your email address in Settings.

12. Supplemental Notice for the EEA, UK, Switzerland and Türkiye

  • Controller: Generative AI Teknoloji Anonim Şirketi, at the address in Section 15, is the controller of your personal information.
  • Legal bases: see Section 4.8. Face checks and face templates involve biometric data processed to uniquely identify a person, a special category of personal data. We process it only with your explicit consent, which you can withdraw as described in Section 7.3.
  • Your rights: you have the rights in Section 7.4, including the right to object at any time to processing based on our legitimate interests and to direct marketing. Under the KVKK, you also have the rights in Article 11, including to learn whether your data is processed, to learn the purpose of processing and the third parties to whom it is transferred, and to request compensation for damage caused by unlawful processing.
  • Automated decisions: our content checks automatically refuse requests and withhold results that appear to break our content rules, and the face check automatically blocks a photo until the people in it are verified. These decisions only affect whether that request is generated or that photo can be used; they have no legal or similarly significant effect on you. For a refused request the app offers to ask a person to review it, unless your account has been blocked for breaking our Terms of Service; for any other decision you can email us at hello@aistudio.works.
  • Complaints: you can complain to the data protection authority where you live or work, such as your EU supervisory authority, the UK Information Commissioner's Office, the Swiss Federal Data Protection and Information Commissioner, or the Turkish Personal Data Protection Authority (KVKK Kurulu). We would appreciate the chance to address your concern first.
  • Providing information: providing your content is necessary to use the generation features; without it we cannot create your result. Face verification is necessary only if you want a real person to appear in your creations. Other information is optional unless we say otherwise.

13. Children

The Service is intended only for adults aged 18 or older, or the age of majority where they live if that is higher. Only adults may use the app and hold an account. We do not knowingly collect personal information from anyone under 18 as a user, and consent links may be confirmed only by adults. If you believe a minor has given us personal information, contact us at hello@aistudio.works and we will delete it.

Adults may upload photos of a child only if they are the child's parent or legal guardian or have that person's permission, and only in line with our Terms of Service. If you want your child to appear in what you create, you, as the parent or legal guardian, verify the child yourself: you take the selfie of the child on your own device and give the consent for the child. Never send a consent link to a child. The child's reference photo and face template are kept on your account under the same rules as any character (Sections 3.4 and 11.4); you can delete them at any time and we delete them on request. If you believe a child under 13 has used the app or confirmed a link, email hello@aistudio.works and we will delete the information.

14. Third-Party Services and Links

The Service may link to or work with services we do not control, such as Apple's App Store, the apps you share your creations to and the websites of our providers. Their own privacy policies govern how they handle your information. When you share an Output outside AI Studio, the recipient service controls what happens to it.

15. Contact Us

If you have questions about this Policy or want to exercise your rights, contact us:

Generative AI Teknoloji Anonim Şirketi (AI Studio)
İdealtepe Mah. Dik Sok. No:13, Daire 2, Maltepe, İstanbul 34841, Türkiye
Email: hello@aistudio.works
In the app: Settings › Need Help › Contact Us