Policy summary
- AI sharing is optional. Content is sent only when you actively use an AI feature and, where required, after you select “Agree and Continue” on the separate AI Data Sharing authorization screen.
- If you select “Not Now,” AI content is not sent. Non-AI, on-device document tools remain available.
- AI requests pass through office.offerapp.ai and OpenRouter, which may route them to an OpenAI or Microsoft Azure model endpoint.
- You can withdraw authorization in Me > AI Data & Privacy. Withdrawal stops new AI transmissions but does not by itself delete previously processed data.
- We do not use document content for advertising, sale of personal information, or cross-app tracking.
Developer, legal entity, and contact
The App is operated by 西安宇之麟网络科技有限公司 (English reference name: Xi'an Yuzhilin Network Technology Co., Ltd.; “we,” “us,” or “our”).
For privacy questions, access requests, or deletion requests, contact:
Documents, local data, and device permissions
Documents you create, open, scan, import, or edit may remain on your device or in storage locations you choose. Viewing, basic editing, conversion, and other non-AI tools do not require us to send document content to an AI provider.
Permissions used only when you request the related feature
- Camera: scan documents or capture pages.
- Photos: import images or save an output you choose.
- Microphone: record dictation, meetings, or voice notes.
- Speech Recognition: convert speech to text. Depending on device capability and settings, audio may be processed on-device or by Apple Speech.
- Files: open, import, edit, export, or save files you select.
You can deny or later change permissions in iOS Settings. Denial affects only the related feature.
AI features and data sharing
We collect and transmit AI content only when you deliberately start the corresponding function. Depending on the feature and your selection, a request may include:
- text, prompts, topics, instructions, drafts, and email content;
- language, tone, length, audience, format, and output settings;
- the file name and content of files you actively select;
- PDF or Word document text and spreadsheet cells;
- ChatPDF PDF content and your questions;
- meeting transcript text, title, transcript segments, action items, and bookmarks;
- an App-specific device identifier where used for request integrity, rate limiting, or abuse prevention; and
- IP address, request time, request path, user-agent information, and security logs generated when the request reaches our server.
AI outputs may include writing, rewrites, email drafts, summaries, extracted information, ChatPDF answers, spreadsheet analysis, presentation outlines, meeting minutes, action items, and bookmarks.
Who receives AI and document data
| Recipient | Data and role |
|---|---|
| Xi'an Yuzhilin Network Technology Co., Ltd. office.offerapp.ai | Receives the request, extracts text where necessary, sends the AI prompt to the gateway, returns the result, and maintains security logs. |
| OpenRouter, Inc. | AI gateway that routes prompts and responses to a model provider. See OpenRouter Privacy Policy and Data Collection. |
| OpenAI, L.L.C. and/or Microsoft Corporation (Azure) | Processes text using an OpenAI model endpoint selected by OpenRouter. See OpenAI Privacy, OpenAI API Data Controls, and Microsoft Privacy. |
| cloud.offerapp.ai, Nextcloud, and Collabora Online | Stores and edits remote documents only when you choose that workflow. Nextcloud stores the file; Collabora processes the live editing session. See Nextcloud Privacy and Collabora Privacy. |
| Apple Inc. — Speech Recognition | May receive speech audio and technical information when Apple Speech performs online recognition. See Apple Privacy. |
We do not authorize recipients to use App content for advertising or cross-app tracking. Providers may process limited metadata for billing, security, fraud prevention, and abuse monitoring under their terms.
Nextcloud and Collabora remote editing
If you choose remote editing, the selected file is uploaded to our Nextcloud instance at cloud.offerapp.ai. Collabora processes the active editing session and saves changes back to Nextcloud. Collabora has no separate persistent document volume in our current production configuration; the persistent copy is held by Nextcloud.
Files remain until you delete them or ask us to delete them. A normal deletion moves a file to the Nextcloud trash bin. You may empty the trash bin to permanently delete it. File versions may exist under Nextcloud's automatic version-management schedule.
Why we process data
- AI writing, rewriting, polishing, shortening, expanding, and continuation;
- email drafting and reply generation;
- document and spreadsheet organization, analysis, extraction, and summarization;
- ChatPDF summaries and question answering;
- presentation outlines and document-based presentations;
- speech-to-text and meeting-minute generation;
- remote storage and collaborative editing when selected;
- service delivery, reliability, security, rate limits, and abuse prevention; and
- support, privacy requests, deletion, and legal compliance.
We do not sell user content or use it for targeted advertising or cross-app tracking.
How long data is kept
| Data category | Current retention or deletion event |
|---|---|
| Local documents and App data | Remain on your device or chosen storage until you delete them, remove the App and its data, or delete them through the storage provider. Device backups may retain a copy under your settings. |
| Ordinary AI prompts and results on our server | Processed in memory for the request. Eligible low-temperature results may be cached in volatile memory for up to 2 minutes. The current service does not write ordinary prompt content to a persistent application database. |
| Uploaded files | The temporary upload is deleted when request processing finishes. Extracted text follows the rule for the feature that uses it. Residual data from an interrupted request is removed through operational cleanup or a deletion request. |
| ChatPDF | The temporary PDF is deleted when processing finishes. Extracted text, summaries, and Q&A history remain in volatile service memory until process restart, administrative cleanup, or completion of a verified deletion request, whichever occurs first. They are not written to the current service's persistent database. |
| AI Excel | The temporary file is deleted when processing finishes. Parsed rows, analysis, charts, and question history remain in volatile service memory until process restart, administrative cleanup, or completion of a verified deletion request. |
| Meeting transcription and minutes | The original transcript is used to generate the response and is not retained in the current persistent application database. Generated minutes, action items, and bookmarks remain in volatile memory until process restart, cleanup, or a verified deletion request. App copies remain until you delete them. |
| Nextcloud / Collabora documents | Active files remain until deletion. Nextcloud's current automatic trash policy normally keeps deleted files for at least 30 days and may keep them longer when space is available, unless you empty trash or request permanent deletion. Operational backups are retired within 90 days, unless law requires longer preservation. |
| IP, device identifier, and web security logs | Nginx logs rotate daily and are kept for up to 14 daily rotations. Diagnostics without document text may be kept longer to investigate a failure or security incident. |
| OpenRouter and downstream AI providers | OpenRouter states content logging is off by default unless an account opts in, while request metadata is retained. OpenRouter may route the current model to OpenAI or Azure. OpenAI states API abuse-monitoring logs may be retained up to 30 days by default; Azure processing follows Microsoft's terms. We do not promise Zero Data Retention. |
| Privacy request records | Correspondence and completion records may be retained for up to 3 years to document compliance, resolve disputes, and prevent unauthorized requests. |
Specific information may be retained longer when required by law, a valid preservation order, dispute resolution, or a security investigation. We limit exceptions to what is necessary.
Consent, refusal, and withdrawal
Before AI data is sent
The App presents a separate AI Data Sharing authorization screen. Data is sent only after you actively select “Agree and Continue.”
If you refuse
Selecting “Not Now” prevents AI content from being sent. Local viewing, editing, file management, and other non-AI tools remain available.
Withdrawal
You may withdraw authorization in Me > AI Data & Privacy. This blocks new AI transmissions unless you later authorize them again.
Access, correction, deletion, and objection
Depending on applicable law, you may ask us to confirm processing, provide access or a copy, correct information, delete information held by us, restrict or object to processing, or explain recipients and purposes.
Email contact@yuzhilintech.com with the subject “AI Office Privacy Request.” Provide only the device or document details necessary to locate the data. We may request reasonable verification to prevent unauthorized deletion.
We will acknowledge and complete a verified request within 30 days, unless applicable law requires a different period. If information was sent to a provider, we will take reasonable steps to support deletion where that provider offers a mechanism.
Security and cross-border processing
We use HTTPS, access restrictions, service isolation, limited production access, log rotation, and deletion controls intended to reduce unauthorized access, alteration, disclosure, or loss. No method is completely secure.
Our servers and providers may process information outside your country, including in the United States or other locations where OpenRouter, OpenAI, Microsoft, Apple, Nextcloud, Collabora, or their infrastructure operates. Laws there may differ from your local law. We limit transfers to what is needed for the requested feature, security, or compliance.
Children's privacy
The App is not directed to children under 13. We do not knowingly collect personal information from a child under 13 without authorization required by law. A parent or guardian may contact us for review and deletion.
Required disclosures and business changes
We may preserve or disclose information when reasonably necessary to comply with law or a valid legal request; protect users, the public, or our rights; investigate fraud, abuse, or a security incident; or establish, exercise, or defend legal claims.
If the App or relevant assets transfer in a merger, acquisition, restructuring, or sale, information may transfer to the successor subject to this policy or notice of a materially different policy.
Changes to this policy
We may update this policy when the App, providers, retention practices, law, or user controls change. The updated version will remain at this URL with a new date. If a change materially expands AI sharing, we will provide in-App notice and request new consent when required.
Privacy contact
西安宇之麟网络科技有限公司
Xi'an Yuzhilin Network Technology Co., Ltd.