Last updated
We hold the account details you give us, the content you create, and the technical data needed to keep your account secure. The prompts and files you send to a model go to the provider that runs it. We do not use your content to train models, and we do not sell your data. Analytics runs only if you accept it. You can delete your chats or your whole account yourself, and email us for anything else.
1. Who is the controller
Omnibo, based in Amsterdam, the Netherlands, is the controller of the personal data described here. Contact: info@omnibo.ai.
This policy covers the Omnibo app (https://app.omnibo.ai), this website and the Omnibo MCP server.
2. The data we hold, why, and on what basis
Under the GDPR every use of personal data needs a legal basis. Ours are: performing our contract with you (“contract”), our legitimate interests in running a secure, working service (“legitimate interests”), your consent (“consent”), and our legal obligations, such as keeping tax and accounting records (“legal obligation”).
| Data | What it includes | Why | Legal basis |
|---|---|---|---|
| Account and profile | Email address, password (stored only as a hash), first and last name, avatar, sign-in method (Google account id if you use Google Sign-In), email verification and password-reset tokens, account status, settings and preferences such as your default model and time zone. | To create and run your account and remember your settings. | Contract |
| Public profile | Username, bio, avatar and follower counts — only if you set up a public profile. | To show your profile and the items you publish. | Contract |
| Sessions and devices | For each signed-in device: device name and type, browser, operating system, IP address, approximate city and country (looked up from the IP address through ip-api.com), and when the session started and was last active. | To keep you signed in, let you see and sign out your devices, and protect your account. | Contract; legitimate interests |
| Chats and AI memory | Your chats and their messages, the files and settings used in them, the tool calls the AI made, and the facts the AI keeps in a project’s memory when you chat inside a project. | To provide chat, keep your history and let the AI work with your projects. | Contract |
| Generated media and files | Images, videos, audio and transcripts you generate; files and reference media you upload; custom voices you create, including the voice recording you upload for them. | To generate, store and deliver your work. | Contract |
| Billing | Your plan and subscription status, the Stripe customer id, purchases, credit balances and usage history, refund requests and payment disputes. Card details are entered on Stripe and never reach our servers. | To bill you, apply credits, handle refunds and keep accounting records. | Contract; legal obligation |
| Newsletter | Whether you subscribed to our newsletter and product news, and when; your email address and name are sent to Mailchimp while you are subscribed. | To send you the newsletter you asked for. | Consent |
| Analytics | Pages visited, clicks on calls to action and, in the app, your account id and a purchase event at checkout — only after you accept analytics cookies. | To understand how Omnibo is used and improve it. | Consent |
| Error monitoring and session replay | Error reports and performance traces from the app and the backend, with known sensitive fields scrubbed; recordings of a sample of app sessions with all text masked and all media blocked. | To find and fix bugs. | Legitimate interests |
| Abuse and security signals | Signals our automated checks record when a chat message, an MCP call or a sign-in matches a known abuse pattern, flags raised for review, links between accounts that sign in from the same network and device, and the decisions taken on them. | To protect the service, its users and its providers from attacks, fraud and abuse. | Legitimate interests |
| Connected accounts | When you connect GitHub or Gmail: the account name, the permissions you granted and the access token, encrypted at rest. | To let the AI use the account you connected, only as you allowed. | Contract |
| MCP keys and authorised apps | The keys you create (we store only a hash and a short prefix), the apps you authorise, and a log of each tool call — which tool, when and the outcome, with the names of its arguments but not their content. | To run the MCP server, show you what your keys did, and investigate misuse. | Contract; legitimate interests |
| Team workspaces | The teams you belong to, your role, invitations, and the content and usage in each team workspace. | To run team workspaces and their billing. | Contract |
| Feedback | The feedback you send from the app: your message, its category, an optional screenshot, the page you were on and your browser’s user agent. | To read and act on your feedback. | Legitimate interests |
| Explore | The items you publish to Explore, and the likes and follows you give and receive. Public views of Explore never show your email address. | To show what you chose to publish. | Contract |
| Emails with us | What you write to us and our replies. | To answer you and resolve your request. | Legitimate interests; contract |
3. Your prompts and the AI providers
When you use a model, your prompt, the conversation context, and any file or reference you attach are sent to the provider that runs that model, which returns the result. The provider processes them under its own terms, which the Model provider terms page names per provider. Some tools send data to other services: a web search sends your query to Tavily, and a connected account is called on your behalf.
We do not use your content to train AI models, and we do not sell your personal data.
4. Analytics and cookies
We use Google Analytics 4 on this website and in the app, but only after you accept analytics cookies. Until then the Google Analytics script is not loaded at all, and with a Global Privacy Control or Do Not Track signal it never runs. If you accepted analytics when you bought a plan or a credit pack, our server also reports that purchase to Google Analytics. You can withdraw your consent at any time; the Cookie Policy lists every cookie and explains how.
5. Error monitoring and session replay
We use Sentry to catch errors in the app and on our servers. Error reports include technical context, such as the page, the browser and the failing request, with known sensitive fields removed before they are sent. Sentry also records about 10% of app sessions, and the sessions in which an error occurs, as a replay of what happened on screen: every piece of text is masked and every image and video is blocked, so a replay shows the layout and the clicks, not your content.
6. Security and abuse prevention
To protect the service we automatically check chat messages, MCP calls and sign-ins for known attack patterns, such as attempts to reach our internal systems, to override the AI’s instructions or to probe for secrets, and we notice when several accounts are created from the same network and device. These checks only record signals and flag an account; a person reviews every flag before anything happens to an account. We also block disposable email addresses at sign-up, limit the rate of sign-in attempts, and use Cloudflare Turnstile to tell people from bots at sign-up, sign-in and password reset.
7. Who we share data with
We share personal data only with the providers that run parts of the service for us (our sub-processors, listed with what each receives), with the AI provider of the model you choose, and with a connected account when you ask the AI to use it. We disclose data to authorities only when the law requires it.
8. International transfers
Our servers, database and file storage run on Railway in its EU West region in Amsterdam, the Netherlands. Several of our providers, including most AI model providers, are based outside the European Economic Area and may process the data they receive there, under the terms named for each of them. When you choose a model, your prompt and inputs go to the provider that runs it.
9. How long we keep data
- Your account and its content are kept until you delete them, or until you delete your account.
- Sign-in: the access cookie lasts 15 minutes and a session 7 days by default, renewed while you use Omnibo. Expired sessions are deleted every day.
- Chats are kept until you delete them, on every plan. A deleted chat can be restored for 7 days and is then permanently deleted, together with its diagnostic records.
- Files no longer linked to anything in your account are removed from storage by a daily cleanup.
- Team workspaces: your data in a team you left is deleted 90 days after you left; all contents of a team whose subscription ended are deleted 90 days after it ended, unless it subscribes again.
- Billing and tax records are kept as long as the law requires.
- The newsletter: until you unsubscribe or delete your account.
10. Staff access
A small number of Omnibo staff with administrator rights can access account data when it is needed to support you, fix a problem or investigate abuse.
- Chat diagnostics: we keep a detailed technical record of each chat turn — the messages, the tools used and the requests to and responses from providers — in private storage. Administrators can download it to investigate a problem; every download is logged, and the record is deleted with the chat.
- “View as”: an administrator can view the app as a specific user to reproduce a problem. Each such session is logged, ends after 30 minutes, and cannot change payments, plans or identity details.
- Every administrator action is recorded in an audit log.
11. How we protect your data
- Sign-in cookies are httpOnly and same-site, and every request that changes data carries a CSRF token.
- Passwords are stored only as salted hashes; MCP keys are stored only as hashes.
- Access tokens for connected accounts are encrypted at rest.
- Generated items are private unless you publish them.
- Sessions and the account status are checked on every request, so signing a device out or suspending an account takes effect at once.
12. Your rights
Under the GDPR you have the right to access, rectify and erase your personal data, to receive it in a portable form, to object to processing based on legitimate interests, to restrict processing, and to withdraw consent at any time. Much of this is self-service:
For anything else — a copy of your data, an objection or a restriction — email info@omnibo.ai from the address on your account. You also have the right to lodge a complaint with a data protection supervisory authority, such as the one where you live.
- Erasure: delete your account under Settings → General → Danger zone. This cancels your subscription, releases the teams you own or belong to, deletes all your files, chats and sessions, your credits, usage history and preferences, and permanently removes you from our newsletter list at Mailchimp. If you own a team with an active subscription, cancel that subscription first.
- Delete all your chats at once under Settings → General → Danger zone, or a single chat from the chat list.
- Access and portability: download your generated and uploaded files from the Files page.
- Rectification: change your name, username and bio under Settings → General → Profile.
- Consent: unsubscribe from the newsletter under Settings → General → Email updates, and change your analytics choice under Settings → General → Privacy & cookies.
- Sign out other devices under Settings → General → Security.
13. Changes to this policy
When we change this policy we update the “Last updated” date above. If a change materially affects how we use your data, we will tell you by email or in the app before it applies.
14. Contact
Questions about this document? Email info@omnibo.ai.