Privacy Policy
Last updated November 11, 2026
The short version
- TIBERO records the office through the microphones you turn on. It transcribes what is said and labels each line as work or personal.
- By default only work talk is shown, searched, turned into tasks or sent to agents. Personal lines are kept out of view. Only your own teammates can open them, on purpose, and partners and guests never can.
- Your workspace belongs to your organization. We don't sell your data, show you ads, or use your content to train AI models.
- We use a short list of providers to run TIBERO, such as Google Cloud for servers and Anthropic and OpenAI for language models. They are listed below.
- You can ask us to see, export or delete your data at any time: [email protected].
1. Who we are
TIBERO is run by Tibero AI ("Tibero", "we", "us"). This policy covers the TIBERO web app at office.tibero.ai, the helper you can install on your computer, and TIBERO's connection for AI agents. It explains what we collect, why, who we share it with and the choices you have.
Each organization's workspace is run for that organization. For recordings, transcripts and other workspace content, your organization decides what is recorded and who can see it, and we process that content on its behalf. Questions about what your organization records are best asked of your organization first. We'll help too.
2. What we collect
Your account
- Your name, email address, and your organization's name and website.
- If you sign in with Google or GitHub, the account ID and verified email address they share with us. We don't get your password, and we don't ask for access to your files or code.
- Your settings, such as which agent you use and which notifications you want.
Office recordings and what comes from them
- Audio from the microphones teammates turn on in their browsers, in short segments while someone is speaking.
- Transcripts of that audio, with the time, the microphone it came from and who was speaking.
- Labels on each line: work, personal, dictation or media, such as a video playing.
- What TIBERO makes from work talk: tasks, daily briefs, answers to questions, and notes about your company, such as its glossary and who does what.
- Transcripts you import from tools like Zoom, Google Meet, Teams or Otter, and images you attach to tasks or messages.
Voice profiles
To label who is speaking, TIBERO builds a voice profile for each person. A voice profile is a set of numbers computed from short clips of your voice, together with some of those clips. It starts when you read the enrollment sentence, or from office recordings once TIBERO matches a voice to you, and it improves from lines you confirm as yours. TIBERO also keeps a voice measurement for each recorded line, so it can tell speakers apart. Voice profiles may count as biometric data in some places. We use them only to label who is speaking in your organization's workspace, and we never sell them or share them outside it.
You can delete your voice samples any time in Profile, under Your voice. TIBERO may still recognize you from what it already learned in the office. To have everything voice-related about you removed, write to us and we'll delete it. We also delete it when you leave your organization, when your organization's workspace is deleted, and in any case within 3 years of your last use of TIBERO.
From the helper on your computer, if you install it
- The computer's name, which agents are installed and connected, recent project folder names, and the apps it can open tasks in.
- For agent sessions: the project, what the agent was asked, which files it touched and whether it is still working. When a teammate opens a session in TIBERO, the helper streams that session's conversation for as long as they're looking.
- If you turn on Wispr Flow matching: the text, time and app of every dictation you make on that computer, and up to 500 words from your Wispr dictionary. TIBERO uses them to recognize your dictations in the room audio, keep them out of transcripts and tasks, and spell your words right.
- When you set up the Agent step: the names of the tools your agent is connected to, such as Google Drive or Slack. We never read what's in them.
Other things you give us
- A phone number, if you choose to get texts. Texts go out through the Messages app on your own Mac.
- If you run agents on a cloud machine: the Anthropic or OpenAI API key you connect. It is stored in Google Cloud Secret Manager, and only its last 4 characters are ever shown.
- Billing details, handled by Stripe. We see your plan and the last digits of your card, never the full card number.
Automatically
- Which pages and features you use, with the time: the pages you open and how long you spend on each (the time you were actively using it, apart from the time it was only left open), the buttons and links you press, and how far you got through signing up and setting up. For a button we keep its label and for a link its address on our site, never anything you type. Before you sign in we know you only by a random id for that browser tab. We use this to fix problems and to understand how TIBERO is used, and it stays with us: no outside analytics company receives it.
- Your IP address and browser details, as any website receives them. We record the IP address of each sign-in, for security.
3. How personal talk is handled
Every line is labelled automatically. Lines labelled personal, such as family, health, errands or weekend plans, are kept out of everything TIBERO shows by default. That includes the transcript, search, tasks, briefs and anything sent to agents. They are still stored in your organization's workspace, along with the audio. Members and admins of your own organization can open them on purpose, from the All tab, by playing a line's audio, or by asking for everything. Partners and guests can never see them, and TIBERO staff never look at them. The labelling is automatic and can make mistakes, so tell your team about TIBERO and turn the microphone off for conversations that shouldn't be recorded.
4. How we use information
- To run TIBERO: transcribe, recognize speakers, label lines, write tasks and briefs, answer questions, and send tasks to the agents you choose.
- To keep your account secure, sign you in and prevent abuse.
- To support you, and to fix and improve TIBERO. TIBERO also improves its instructions for your workspace from your team's own corrections. That stays in your workspace and is never used for another organization.
- To bill your organization, and to send sign-in links, invites and messages about your account.
We don't sell your personal information, use it for advertising, or use your content to train AI models. Where the law asks for a legal basis, we rely on running our contract with you or your organization, our legitimate interest in keeping TIBERO working and secure, and your consent for voice profiles and texts.
5. Who we share it with
Your organization. Your teammates in the same workspace see work talk and tasks. Admins manage members and settings.
Service providers that run parts of TIBERO for us, only for that purpose:
- Google Cloud: the servers and storage TIBERO runs on, in the United States, and secret storage.
- Cloudflare: the network between your browser and our servers.
- Anthropic and OpenAI: language models that clean up transcripts, label lines, write tasks and briefs, and answer questions. They receive the text needed for each request. Cleanup and labelling see every line, including the ones that turn out to be personal, because sorting them is the job. Their API terms don't allow them to train their models on it.
- Stripe: payments.
- Google and GitHub: signing in, when you choose them.
- The email provider that sends sign-in links and invites.
Speech-to-text and voice recognition run on our own servers. Your audio isn't sent to an outside speech service.
Agents you choose. When you send a task to an agent, such as Claude Code, Codex or the Claude app, the task and its context go to that agent under your own account with its provider. That provider's terms and privacy policy apply.
AI apps you connect. When you connect TIBERO to an AI app, such as Claude, ChatGPT, Claude Code or Codex, and sign in, that app can read what you can read in TIBERO, as you: work transcripts with who said what and when, conversation notes, tasks, notes and your company's memory, and it can create and update tasks and notes for you. It reads only when you ask it something, and personal talk stays out unless you ask for it. What it reads goes to that app's provider under your own account, and its terms and privacy policy apply. You can disconnect it at any time in Profile or in the app, and removing someone from your workspace disconnects their apps.
TIBERO staff. A few people at TIBERO can see each organization's usage counts. To support you, they can also open your task board and work transcripts read-only. Every look is recorded in your workspace's activity, and your admin can turn this access off in Profile. Staff never open personal lines or audio. Because your data lives on servers we run, this is our rule rather than a technical barrier: the few engineers who run the servers could reach it, and they do so only to keep TIBERO working.
The law and changes of ownership. We share information when the law requires it, or to protect people's safety or our rights. If Tibero is merged or sold, your information would move with it under this policy, and we'll tell you first.
6. How long we keep it
- Recordings, transcripts, tasks and other workspace content stay while your organization uses TIBERO. Some things, like tasks, imports and voice samples, you can delete yourself; for everything else, your admin can ask us to.
- Voice profiles: until deleted, as described above.
- Sign-in links expire after 15 minutes, and sign-in sessions after 180 days.
- When we receive a verified request to delete an account or a workspace, we delete it within 30 days, apart from what the law requires us to keep, such as billing records.
7. Security
Everything between your browser and Tibero is encrypted in transit. Stored data is encrypted at rest by Google Cloud. Each organization's transcripts and recordings are stored on a private virtual machine (VM) that is unique to that organization, separate from every other organization's. Secrets such as API keys live in Google Cloud Secret Manager. Only a small number of Tibero engineers can reach the servers, and only to run and fix the service. No system is perfectly secure, so please tell us at once if you think your account has been misused.
8. Your choices and rights
- Microphone. You decide when your microphone listens, and you can turn it off at any time.
- Voice profile. Delete it any time in Profile, under Your voice.
- See, correct, export or delete. Write to us and we'll help, usually within 30 days. If your data belongs to an organization's workspace, we may check with its admin first.
- Usage records. Turn on Global Privacy Control or Do Not Track in your browser and TIBERO keeps no record of the pages you open or the buttons you press.
- Object or withdraw consent. You can withdraw consent for your voice profile or texts at any time. It doesn't affect what we did before.
- Where you live. Depending on where you are, for example in the EU, the UK or California, you may have more rights. These include knowing what we hold, asking us to delete it, and complaining to your data protection authority. We'll never treat you differently for using them.
People who aren't TIBERO users but were recorded in an office can also write to us. We'll work with that organization to answer.
9. Cookies and browser storage
We use two cookies, both needed to sign in. One keeps you signed in for up to 180 days, or 30 days when you open a sign-in link. The other lasts 10 minutes, only while a Google or GitHub sign-in finishes. Your browser also stores small settings, like where you are in the onboarding, and a random id for the tab you are in, which ends when you close it, so one visit can be told from another. We don't use advertising or tracking cookies, and we don't run third-party analytics. Some pages load fonts from Google Fonts, so your browser asks Google for them.
10. Children
TIBERO is for work and isn't meant for anyone under 16. We don't knowingly collect children's information. If you think we have, tell us and we'll delete it.
11. Where data is stored
TIBERO's servers are in the United States. If you use TIBERO from elsewhere, your information is transferred there. Where the law requires it, we use appropriate safeguards for those transfers, such as standard contractual clauses.
12. Changes to this policy
We'll post any update here and change the date at the top. If a change matters, we'll also tell admins by email or in the app before it applies.
13. Contact
Questions or requests: [email protected]. See also our Terms of Service.