1AI (the "App") is an AI assistant. Its core function is to take what you write and send it to an artificial-intelligence model so that model can answer you. Those models are built and run by other companies. This Privacy Policy explains exactly what leaves your device, who receives it, what I do and do not do with it, and the choices you have. Please read it carefully. By installing or using any part of the App or this website, you agree to it.

1AI is built and operated by one person. "I" and "me" below refer to Chat Labs, the sole developer of the App; "you" refers to the person using it.

Key Highlights

The App sends your chat content to third-party AI providers. Messages, attached files, images, voice recordings and web page addresses you ask the App to read are transmitted off your device so an AI model can respond. Section 3 names every recipient.
The App asks before it does this. On first launch you are shown a disclosure describing what is sent and to whom, and nothing is transmitted for AI processing until you accept it and send a message.
I do not sell your information — not to anyone, for any purpose.
I do not use your conversations to train AI models, and I request the same of the providers I route through (Section 5).
Your conversation history is stored on your device, not on a server I control, unless you deliberately create a share link (Section 2.2).
I collect the minimum information necessary to run and improve the App, and I use analytics that do not identify you.
All network traffic is transported over HTTPS.

1. Information I Collect And How I Collect It

I use the term "Consumer Information" to mean any personally identifiable or anonymous information relating to you or your use of the App or website. Information reaches me in three ways: you type or attach it in the App, the App generates it automatically as you use it, or Apple provides it to me in connection with a purchase.

1.1 Chat Content

Chat Content is everything you put into a conversation: the text of your messages, files you attach (such as PDF, TXT, MD and CSV documents), images you attach or capture, voice recordings you dictate or speak in a voice conversation, and the addresses of web pages you ask the App to read or search for. Chat Content also includes any custom instructions, saved prompts or agent-memory notes you configure, because those are sent alongside your message so the model can follow them.

Chat Content is collected because you typed it, and it is transmitted because that is the only way an AI model can answer. See Section 3 for who receives it and Section 4 for how long it is kept.

1.2 Contact Information

Contact Information is information that identifies or lets someone contact you, such as your name or email address. The App has no account system and no sign-in: I do not intentionally collect Contact Information, and if you do not contact me, your use of the App is anonymous to me. If you email support or submit feedback, whatever you include in that message is used solely to answer you.

1.3 Usage Information And Analytics

Usage Information covers how the App is used in aggregate: which screens are opened, which features are used, how many people use the App per day, and in which countries, languages, devices and OS versions. This is collected through a self-hosted, cookieless Umami instance that I run on my own infrastructure; it is not shared with an advertising network and it is not linked to your identity or to your Chat Content. Analytics events record that a feature was used — never the text of what you wrote.

1.4 Diagnostic And Crash Information

When the App crashes or hits an error, a report is sent to a third-party crash-reporting service containing the error, a stack trace, the App version, the device model and OS version. These reports are used to find and fix bugs. Crash reports are not deliberately populated with Chat Content; where an error message unavoidably includes a fragment of a request, it is retained only for the diagnostic period in Section 4.

1.5 Purchase Information

Purchase Information is the history of your subscription — when you subscribed, renewed, cancelled or were refunded. Purchases are processed by Apple, and I never see your payment method or card details. Subscription status is managed through RevenueCat, which receives a pseudonymous identifier for your installation and the transaction data Apple provides, so the App knows whether your subscription is active.

1.6 Device Token And Scheduled Tasks

If you enable notifications or scheduled tasks (such as a morning briefing), the App sends the Apple push token for your device, plus the schedule and the instruction you configured, to my notification backend so the task can run at the time you asked and the result can be delivered to you. Running a scheduled task means sending its instruction to an AI provider on your behalf, exactly as Section 3 describes.

2. How I Use And Store Your Information

2.1 Purposes

I use Consumer Information to operate the App and answer your requests; to provide customer support; to process and restore subscriptions and detect purchase abuse; to deliver notifications and scheduled tasks you configured; to diagnose crashes and fix bugs; to understand in aggregate which features are used so I know what to build and which languages to translate into; and to comply with legal obligations. I do not use it for advertising, and I do not sell it.

2.2 Where Your Conversations Live

Your conversation history is stored in a database on your device. I do not run a server that holds a copy of your chats, and there is no cloud sync. Two things change this, and both require a deliberate action by you: creating a share link uploads that one conversation so the link can be opened (share links expire automatically after 7 days), and sending a message necessarily transmits that message to an AI provider as described in Section 3.

Deleting a conversation in the App deletes it from your device. It does not, and cannot, reach back and delete the copy the AI provider received when you sent it; that copy is governed by the provider's own retention policy.

2.3 The Relay

Requests from the App do not usually go straight to the AI provider. They pass through a relay service I operate, which holds the API credentials, applies rate limits and routes each request to the correct provider. The relay runs on third-party cloud hosting infrastructure. It processes your request in transit; it is not a conversation archive. Two features connect directly from your device to the provider instead of through the relay, and they are marked as such in Section 3.

3. AI Providers And Other Third Parties Who Receive Your Data

This is the section that matters most. When you send a message, the content described in Section 1.1 is transmitted to the company that operates the AI model you selected. Which company that is depends on which model you pick in the App.

3.1 AI Model Providers

OpenAI — GPT models, image generation, text-to-speech, and the live voice conversation feature. Voice conversations connect from your device directly to OpenAI.
Anthropic — Claude models.
OpenRouter — a routing service that forwards your request to whichever company builds the model you chose. Through OpenRouter the App offers models from Google, Meta, DeepSeek, Mistral, xAI, Alibaba (Qwen), Cohere, Amazon, Perplexity, Microsoft, MiniMax, Moonshot, ByteDance, Tencent, NVIDIA, Z.ai and others. Selecting one of those models means your Chat Content reaches both OpenRouter and that company, each under its own privacy policy.
Replicate — image generation and image editing models.
Deepgram — speech-to-text. Audio you dictate or record for transcription is sent to Deepgram, or to OpenAI, depending on the transcription mode in use.
Fish Audio — text-to-speech for spoken output in some voices.
Exa and Serper — web search. When you use a search or research feature, your query is sent to a search provider.
Jina AI — web page reading. When you ask the App to read a page, that page's address is sent from your device directly to Jina AI, which fetches and returns the page text.

If you configure the App to use a local or self-hosted model through Ollama, your Chat Content goes to the server you specified and to no one else. That is the one configuration in which no third-party AI provider is involved.

Each provider handles the data it receives under its own privacy policy, linked above. I select providers that offer business or API terms committing not to train their models on API content, and I do not grant any provider permission to do so. I cannot, however, guarantee the internal practices of a company I do not operate — which is why the App tells you who receives your data and asks you before sending it, so the choice is yours to make.

3.2 Service Providers

Beyond AI providers, a small number of services support the App: Apple (App Store purchases and push delivery), RevenueCat (subscription status), a cloud hosting provider (infrastructure for the relay and the notification backend), a crash-reporting service (diagnostics), my self-hosted Umami instance (analytics), and a form service used for optional feedback submissions. Each receives only what it needs for its function, and each is engaged under terms that require it to protect that information and to use it only to provide the service to me — not for its own purposes. I will name any of them on request, at the address in Section 10.3.

3.3 What I Do Not Do

I do not sell your Consumer Information. I do not share it with advertisers or data brokers. I do not disclose your Chat Content to any third party other than the providers named above that are required to answer your request, or as compelled by law under Section 7.

4. How Long Information Is Kept

Conversations: on your device, until you delete them or delete the App.
Share links: automatically disabled 7 days after creation.
Requests in the relay: held only for the duration of the request. Operational logs record that a request occurred and its size and cost, not its content.
Data held by AI providers: governed by each provider's policy. Most retain API requests for a limited abuse-monitoring window (commonly up to 30 days) and then delete them. Consult the policies linked in Section 3.1 for the current terms.
Analytics: aggregate, non-identifying, retained indefinitely in aggregate form.
Crash reports: up to 90 days.
Purchase records: retained as long as required for subscription management, refunds and tax obligations.
Support emails: until you ask me to delete them (Section 6).

5. Model Training

I do not use your conversations, files, images or voice recordings to train, fine-tune or evaluate any AI model. I do not build datasets from user content. Where a provider offers a training opt-out or API terms that exclude training by default, I use them.

6. Your Choices And Controls

Consent: the disclosure shown on first launch is the point at which you decide. You can re-read it any time from Settings.
Choosing a provider: picking a different model changes which company receives your data. A local model through Ollama sends it to no third party at all.
Deleting conversations: delete individual chats or all of them from within the App; deleting the App removes all of it from your device.
Memory and personalization: agent memory and saved instructions can be turned off in Settings, which stops them being attached to your requests.
Notifications and scheduled tasks: can be declined or turned off at any time, in the App or in iOS Settings.
Microphone, camera and photos: the App can only use them with the permission iOS asks you for, and you can revoke it in iOS Settings.
Access and deletion: email me and I will delete the support correspondence and any diagnostic data I hold that is associated with you.

7. Security And Legal Disclosure

All network traffic between the App, my relay and the providers is encrypted in transit over HTTPS, and the App additionally pins the certificate of its own relay. I use commercially reasonable measures to protect information I hold, but no method of transmission or storage is 100% secure, and I cannot guarantee absolute security.

I may be legally obligated to disclose Consumer Information in response to a lawful demand — a subpoena, a court or regulatory order, or a comparable legal process — or where necessary to comply with applicable law, to enforce the License Agreement, or to protect the rights and safety of users or the public.

8. International Transfers

The providers and services described above operate infrastructure in the United States and other countries. Using the App means your information may be processed in a country other than the one you live in, including countries whose data-protection laws differ from your own.

9. Children

The App is not directed at children under 13, and I do not knowingly collect information from them. If you believe a child has provided information through the App, contact me and I will delete it.

10. General

10.1 Changes To This Policy

I may update this Privacy Policy. When a change materially affects what is sent or who receives it, the App will present the disclosure again so you can decide with the current facts in front of you. The date below always reflects the latest revision.

10.2 Technical Problems

Despite the measures described here, technical mistakes are possible. If one occurs, I will take commercially reasonable steps to correct it.

10.3 Contact

For questions, suggestions, access or deletion requests, email support@chatlabsai.com.

10.4 Interpreting This Policy

Examples are provided to aid understanding and are not exhaustive lists. Section headings are for convenience. This Privacy Policy supplements and is incorporated into the License Agreement.

Last Updated: August 25, 2026