Privacy Policy for WitSpeak

WitSpeak is operated by ZenithNexus AI Technologies LLC, a company based in Miami, Florida, USA (“WitSpeak”, “we”, “us”). We are the data controller for the personal information described in this policy. If you have any questions or requests about this Privacy Policy or your data, contact us at support@witspeak.ai.

• Account information: name, email address, sign-in method (email, Apple, or Google), profile photo if you add one, and basic profile details such as your native language, learning goals, and target scores.
• Acceptance records: the date and version of the Terms, this Privacy Policy, and the voice-data disclosure in effect when you create your account.
• School enrollment details: if you join a class with a school code, we also collect your birth year (to determine whether parental consent is required) and, where consent is needed, a parent or guardian email address.
• Practice and learning data: spoken responses, transcripts, written answers, AI feedback and score estimates, pronunciation results, vocabulary progress, streaks, XP, and other learning activity.
• Voice and audio data: recordings of your spoken practice, processed as described in the “Voice and speaking data” section below.
• Device and usage data: device type, operating system, app version, language and region settings, crash logs, and in-app interaction events.
• Purchase data: your subscription or purchase status from the Apple App Store or Google Play. We never receive your full card number.
• Support communications: messages you send to our support team and the information you choose to include in them.

• Provide speaking practice, conversations, and AI-powered feedback.
• Generate analysis, score estimates, rewrites, and personalized improvement suggestions.
• Personalize your learning plan, difficulty, and content to your level and goals.
• Maintain your account, sync progress across devices, and reflect your subscription status.
• Send service messages, and — if you opt in — practice reminders and product updates.
• Improve app performance, reliability, and the quality of our teaching content.
• Respond to support requests.
• Prevent abuse, fraud, and technical misuse, and enforce usage limits.
• Comply with legal obligations.

Voice practice is the core of WitSpeak. As we disclose when you create your account: when you use speaking features, WitSpeak records, analyzes, and stores your voice and speech-analysis data to provide feedback, track your progress, and improve its pronunciation models. Recording happens only while you are actively using a speaking feature — never in the background.

What happens to the audio itself depends on the feature:

• Live conversation practice (roleplays, daily conversations, and similar real-time speaking): your voice is streamed and processed in real time to run the conversation. We do not keep the audio afterwards — what remains is the written transcript, feedback, and scores.
• Pronunciation, shadowing, and similar speaking exercises: your recording is analyzed to score your pronunciation, and we keep some of these practice clips — together with the practice scores and your native language — to evaluate and improve WitSpeak’s own pronunciation and accent-feedback models, unless you turn this off in Settings (see “Your controls” below). The scores and progress are always kept.
• Speaking test simulations (for example IELTS-style speaking tests): your answer recordings are stored in private storage so you can replay and review them. They are visible only to you (and to our support staff if you ask for help), and they are deleted when you delete them or your account.

Your controls. In the app’s settings you can turn off the use of your practice clips for model improvement at any time — this stops new collection immediately — and you can delete your stored voice recordings at any time, from Settings, by deleting your account, or by emailing support@witspeak.ai. Clips kept for model improvement are stored under a pseudonymous identifier rather than your name or email. If you previously chose not to share your recordings, that choice is preserved — we never re-enable it for you. Accounts connected to a school program (section 12) are always excluded: we never use school students’ practice audio to improve our models.

Audio is processed by the AI and speech providers listed in section 6 under agreements that restrict them from using your data to train their general-purpose models, and a small number of authorized WitSpeak reviewers may access limited samples for quality assurance under confidentiality obligations. We never use your voice recordings for advertising, and we never sell them. We do not create voiceprints and we do not use your voice to biometrically identify you. Voice recordings are never shared with schools or employers — see section 12.

WitSpeak feedback is educational. We do not present it as an official exam score, and it may not match official test results.

WitSpeak uses artificial intelligence to hold conversations, analyze responses, generate feedback and rewrites, and personalize practice. In line with EU AI-transparency rules (including Article 50 of the EU AI Act), we make this explicit: when you practice in WitSpeak, you are talking to an AI system, not a human. The coaches, examiners, and characters in the app are AI-generated, and their voices and videos are synthetic. Feedback, scores, and study suggestions are AI-generated content.

WitSpeak does not use AI to infer your emotions for any purpose other than helping you practice, does not use biometric categorization, and does not make legal or similarly significant automated decisions about you. AI output can be imperfect or incomplete and should be treated as educational guidance, not as a professional evaluation. If you have questions about how the AI works, or want a human to review something, contact support@witspeak.ai — a real person reads those messages.

We share information with a small number of providers that help us run the Service. Each receives only what it needs to perform its function on our behalf:

• OpenAI — AI conversation, feedback, and scoring. Receives practice audio and transcripts.
• Microsoft Azure Speech — speech recognition and pronunciation assessment. Receives practice audio.
• ElevenLabs — speech-to-text for your spoken answers and generation of the coach and examiner voices you hear. Receives practice audio and lesson text.
• Apple App Store, Google Play, and RevenueCat — process purchases and manage subscription status.
• Replit and Neon — application hosting and database infrastructure where your account and progress are stored.
• Cloudflare — content delivery, security, and storage of media such as profile photos and practice audio.
• Resend and SendGrid — deliver account and support emails.
• Expo — delivers push notifications if you enable them.

These providers are not permitted to use your information for their own marketing. Each one processes your data under contractual terms that require it to protect your information with safeguards that provide the same or equal protection as this policy, and to use it only to provide its service to WitSpeak. We may update this list as the Service evolves; the current version is always available on this page.

Purchases are processed by Apple or Google, and subscription records are managed through RevenueCat. We receive confirmation of your purchase and subscription status so we can unlock your plan, but we never receive your full payment card details.

We use analytics to understand how the Service is used and to make it better:

• Amplitude — product analytics and session replays of in-app interactions that help us diagnose usability problems.
• Google Firebase — app analytics and crash reporting.
• PostHog — website analytics, loaded only if you accept the cookie banner on our website.
• AppsFlyer — mobile install attribution, so we know which campaigns bring new learners.
• Meta, TikTok, Reddit, and Google — receive limited conversion events (for example, that a sign-up or purchase happened) so we can measure our advertising. They do not receive your practice content.

Our website uses essential cookies and local storage (for example, to remember your language). Optional analytics cookies are set only with your consent. We do not sell your personal information. Where US state privacy laws treat advertising measurement as “sharing”, you have the right to opt out: use your device’s tracking controls (such as iOS App Tracking Transparency), the website cookie banner, or email support@witspeak.ai and we will honor your request.

We keep your information while your account is active so your progress and history are available to you. In outline:

• Live conversation audio: not retained — processed in real time and discarded (see section 4).
• Pronunciation and shadowing practice audio: clips kept for model improvement (section 4) are retained until you delete them in Settings, delete your account, or ask us to remove them. If you turn model improvement off, new recordings are not kept after scoring.
• Speaking test recordings: kept in private storage until you delete them or your account.
• Transcripts, feedback, scores, and learning progress: kept while your account is active.
• Account and profile data: kept while your account is active.
• Purchase records: kept as long as required for legal, tax, and accounting purposes.
• Analytics and crash data: kept under the retention settings of the analytics providers in section 8, then deleted or aggregated.

You can delete your account at any time in the app’s profile settings or by emailing support@witspeak.ai. When you delete your account, your personal information — including stored recordings and transcripts — is deleted from our active systems, except for limited records we must keep for legal, security, or accounting reasons (such as purchase records). Residual copies may persist briefly in encrypted backups and logs until those are purged on their regular schedule.

Depending on where you live — including under the EU/UK GDPR, Turkey’s KVKK (Law No. 6698), and US state privacy laws such as the CCPA — you have the right to:

• Access the personal information we hold about you and receive a copy of it.
• Correct inaccurate information.
• Delete your account and personal information.
• Object to or restrict certain processing, and withdraw consent where processing is based on consent.
• Manage subscriptions and payments through your Apple or Google account settings.
• Disable microphone access or push notifications in your device settings at any time.
• Complain to your local data protection authority (for example your EU supervisory authority, the UK ICO, or the Turkish Personal Data Protection Authority, KVKK).

To exercise any of these rights, use the in-app options or contact us at support@witspeak.ai. We respond to all requests and will never treat you differently for exercising them.

Age rules for WitSpeak depend on how the account is used:

• Individual (consumer) accounts: you must be at least 13 years old, and at least the digital-consent age of your country if it is higher (this age varies by country — for example 16 in Germany and Slovakia, and up to 16 across much of the EU). Under Turkey’s KVKK, minors need the consent of a parent or guardian.
• School accounts: younger learners (including children under 13) may use WitSpeak only through a WitSpeak for Schools program, with the consent of a parent or guardian.

When a student joins a class with a school code, we ask for their birth year. If the student is below the digital-consent age that applies to them, the account is placed in a pending state and is not activated in the class until parental consent is confirmed — either by a consent link we email to the parent or guardian, or by the school confirming that it holds written parental consent on file. Parents and guardians can review, withdraw consent for, or request deletion of their child’s data at any time via the school or by emailing support@witspeak.ai.

For students in the United States, schools may authorize the collection of student personal information under COPPA solely for educational purposes. Consistent with COPPA and student-privacy rules: we collect from child accounts only what the Service needs, we do not show ads in WitSpeak or use children’s data for advertising or marketing, we do not sell it, we do not build profiles for any purpose other than the learning experience, and we retain it only as long as needed for the educational purpose (section 9). Voice audio from live practice is processed in real time and not retained (section 4), and school accounts are never included in voice-data model improvement.

Outside school programs, we do not knowingly collect personal information from children below the ages above. If we learn that we have done so without appropriate consent, we will delete it. If you believe a child has provided us personal information, contact us at support@witspeak.ai.

If your account is connected to a school through a WitSpeak for Schools program, your teachers (and school staff who administer the program) can see, for the classes you joined:

• Progress metrics: practice minutes, weekly activity, streaks, level progression, scores and score estimates, and homework completion.
• Assignments and announcements: what was assigned to you and whether you completed it.
• Written transcripts of your practice sessions, and the coach's written corrections — only if you agreed to this when joining the class, and only for practice done after you joined. Sessions you mark private are never included, and you can turn transcript sharing off in the app at any time.

Teachers and schools never receive your voice recordings — there is no feature that plays a student’s audio to a teacher. Practice you did before joining the class is not shared, and if you leave the class, sharing stops from that moment. School-linked learning data is used only to provide the educational service to you and your school — never for advertising, and it is never sold. School-linked accounts are always excluded from voice-data model improvement (section 4): we never use school students’ practice audio to train our models.

For students below the digital-consent age, class membership is activated only after parental consent (section 11). We rely on the school to have the authority to run the program, and where the school confirms it holds written parental consent, we record that confirmation. Schools can request deletion of their students’ data at any time.

We operate globally, and your information is processed and stored in the United States and other countries where our providers operate. Where required, we rely on appropriate safeguards for these transfers, such as standard contractual clauses.

We protect your information with technical and organizational safeguards, including encryption in transit and access controls that limit who can reach personal data. No method of transmission or storage is completely secure, so we cannot guarantee absolute security — but if we become aware of a breach affecting your data, we will notify you as required by law.

We may update this Privacy Policy from time to time. When we do, we will revise the effective date at the top of this page, and for material changes we will give you reasonable advance notice (for example, in the app or by email). Continued use of WitSpeak after changes take effect means you accept the updated policy.

ZenithNexus AI Technologies LLC · Miami, Florida, USA

support@witspeak.ai