Privacy
GhostSpeech Privacy Policy
Last updated: July 22, 2026
GhostSpeech is a dictation keyboard for iPhone. You hold or tap to dictate, ask questions, and rewrite text. This policy explains what data GhostSpeech collects, why we use it, when it leaves your device, and the choices available to you.
The short version
- Your audio is transcribed and discarded. GhostSpeech does not store voice recordings.
- The keyboard has no internet access of its own. Network requests are made by the GhostSpeech app only when you invoke a connected feature.
- Normal typing is never logged, recorded, or sent. Only content needed for an action you explicitly invoke is processed.
- You control your history. You can delete items, clear history, change retention, or delete your account from inside the app.
- Private Mode, on supported iPhones, keeps supported speech and language processing on your device.
- We do not sell personal information or use it for cross-app advertising or tracking.
Scope
This policy applies to the GhostSpeech app, keyboard, website, and related services operated by VESTAPLAN, INC. It does not govern services that you choose to access outside GhostSpeech.
Data we collect and why
- Account data: an account identifier and, when provided during Sign in with Apple, an email address and complete name. We use the account identifier and email to create, secure, and support your account. The name and email are also shown as read-only account information inside GhostSpeech; the name is not used to authenticate or authorize the account.
- Audio and user content: audio you submit for dictation or Ask, prompts, transcripts, text you submit for Rewrite, and limited text near the cursor when needed for an Ask request. We process this content only to perform the feature you request, provide history you enable, prevent abuse, and improve the service when you separately opt in.
- Optional location: when you enable location for Ask, approximate location rounded to reduce precision. We use it only to make the requested answer more relevant.
- Subscription data: product, status, and transaction identifiers associated with a pseudonymous account identifier. We use this to validate purchases and provide paid features. We do not receive payment-card details.
- Usage and diagnostics: feature interactions, device and app information, performance timings, and error types. We use this to operate, secure, troubleshoot, and improve GhostSpeech. Diagnostics do not include your speech, typed text, prompts, or transcripts.
- Support communications: messages and information you send when requesting help. We use them to respond and resolve your request.
We receive data from you, your device, the platform services you use to sign in or purchase a subscription, and service providers acting on our behalf.
Dictation and audio handling
When you dictate, audio is sent to our server or, in Private Mode, processed entirely on your device. It is converted to text and then discarded. GhostSpeech does not store voice recordings. The microphone runs in the GhostSpeech app, not in the keyboard; the keyboard borrows it.
Your history and improvement choices
- Ask history stays on your phone. Ask conversations are stored locally on your device only.
- Dictation and rewrite history may be saved to your account so you can review it across sessions. You control retention: Off, last 10, or All. You can delete an item or clear a category. Setting retention to Off also removes existing server history in that category.
- Help improve GhostSpeech is optional. If you opt in under Settings > Privacy, we may retain transcripts and rewrites, text only and never audio, to improve recognition quality even when history is off. This is off by default and can be changed at any time.
What connected features send
Connected Dictate, Ask, Rewrite, search, and media features send only the information needed to perform the action you choose. An Ask request may include text near your cursor, up to about 4,000 characters, and, only if you enabled the Ask location option, reduced-precision approximate location. GIF search sends your GIF query, a provider-scoped random identifier, and the interaction signals required to operate that feature to the media-search provider. Review surrounding text before invoking a connected feature if it may contain sensitive information.
Private Mode
On supported iPhones, Private Mode runs Dictate, Ask, and Rewrite on your device. While it is active, those features do not send your speech, nearby text, prompts, transcripts, or generated content to our servers; media search is disabled; and diagnostics upload is paused. Some features, including web search and location-aware answers, are unavailable in Private Mode.
The keyboard itself
The keyboard extension makes no network connections of its own. It communicates only with the GhostSpeech app on your device, which performs requests for connected features. The keyboard cannot see your screen or other apps, only the active text field. As iOS requires for every keyboard, secure fields such as passwords use the system keyboard instead.
How we disclose data
We disclose the minimum information needed to service providers that perform cloud speech recognition, generative-AI processing, search and media retrieval, hosting, diagnostics, authentication, and subscription processing on our behalf. Their access depends on the feature you choose. They are permitted to process data only to provide and secure the service for us, not to sell your personal information, use it for cross-app tracking, or train general-purpose models on your content.
We may also disclose information when reasonably necessary to comply with law, protect the rights and safety of users or others, investigate abuse, or complete a merger, financing, acquisition, reorganization, or sale of assets. If ownership changes, this policy will continue to apply to transferred personal information unless you are notified otherwise.
Legal bases where applicable
Where data-protection law requires a legal basis, we process information as needed to perform our contract with you and provide requested features; for legitimate interests such as securing, supporting, and improving GhostSpeech, balanced against your rights; with your consent for optional processing; and to comply with legal obligations. You may withdraw consent at any time, but withdrawal does not affect processing already completed.
Retention and deletion
- Source audio is processed transiently and is not retained as a voice recording.
- Account content is retained according to the history and improvement choices described above and until you delete it or your account.
- Diagnostics are pseudonymous while associated with an active account, are de-identified when the account is deleted, and are removed within 180 days.
- After account deletion, a content-free, one-way allowance identifier and its credit ledger may remain for up to 60 days to prevent an immediate welcome-credit reset. It does not contain your Apple identifier, name, email, audio, prompts, transcripts, or typed text and is deleted after that period.
- Account, subscription, security, and support records are retained only as long as needed to provide the service, meet legal obligations, resolve disputes, and prevent fraud or abuse.
Deleting your account removes your account and account-linked GhostSpeech content from our primary systems, disconnects the associated platform sign-in authorization, and removes the subscription-management customer profile, subject to the limited 60-day allowance record described above. Other limited records may be retained where required by law or for legitimate security and fraud-prevention purposes, and de-identified data may be retained where it can no longer reasonably be linked to you. Deleting your GhostSpeech account does not cancel an App Store subscription.
Security and international processing
We use administrative, technical, and organizational safeguards designed to protect personal information, including encryption in transit, access controls, and restricted production credentials. No system is completely secure. GhostSpeech and its service providers primarily process data in the United States, which may have different data-protection laws from where you live.
Children
GhostSpeech is not directed to children under 13. If you are not old enough to enter these Terms or consent to data processing under the law where you live, a parent or legal guardian must consent on your behalf where the law permits. Contact us if you believe a child provided personal information without appropriate consent.
Your choices and rights
- Configure history retention, delete items, or clear a category in Settings.
- Turn diagnostics off or change the Help improve GhostSpeech choice in Settings > Privacy.
- Use Private Mode on supported devices for available on-device features.
- Delete your account and account-linked data from inside the app.
Depending on where you live, you may have rights to access, correct, delete, or obtain a copy of personal information, or to object to or restrict certain processing. We may need to verify your identity before completing a request. To exercise a right or appeal a decision, email privacy@ghostspeech.ai. We will not discriminate against you for exercising a privacy right.
Changes to this policy
We may update this policy as GhostSpeech evolves. We will post the revised policy here and update the date above. If a change materially affects your rights or how we use personal information, we will provide additional notice appropriate to the change.
Contact
VESTAPLAN, INC. - privacy@ghostspeech.ai