Privacy policy

What we collect, why, and where it lives.

This policy covers the Knowledge Commons hosted service at knowledgecommons.ai, the Knowledge Commons companion apps for iOS, Android and macOS, and the Knowledge Commons Web Clipper for Chrome, Edge, Firefox and Safari. It is written to be read, not scrolled past. Effective 2 November 2026.

Who we are

Goal17, Inc. is the operator of the hosted service.

Knowledge Commons is built and operated by Goal17, Inc., Toronto, Canada. For the hosted service at knowledgecommons.ai, Goal17 is the organisation responsible for your personal information under Canada’s Personal Information Protection and Electronic Documents Act (PIPEDA) and comparable laws elsewhere.

Questions, access requests and complaints go to aaron@goal17.co. We answer within 30 days.

Self-hosted deployments are different. When an organisation runs Knowledge Commons on its own infrastructure — air-gapped or otherwise — that organisation controls every byte and is the responsible party. Goal17 receives nothing from a self-hosted deployment. This policy governs only the hosted service and the apps when they are pointed at it.

What we collect

Only what the product needs to work.

Everything below is collected because a feature depends on it. There is no advertising, no profiling, no third-party analytics SDK and no sale of data — in the apps or on this site.

  • Account information — Your name, email address and a user identifier, created when your organisation invites you or when you sign up. Used to sign you in, to show who contributed what, and to send the emails that operate the service — invitations, verification, billing notices.
  • Audio recordings — Recordings you make with a companion app, a Knowledge Commons hardware recorder or the desktop recorder are uploaded to your organisation’s project and transcribed. Recordings may include the voices of other people in the room. You are responsible for obtaining any consent the law requires before recording. The platform’s Anonymous mode strips speaker identity from the transcript; Speaker-ID mode keeps it. Your project’s administrator chooses which.
  • Documents, notes, photos and other content — Files you upload, notes you write, pages you scan with the app’s camera, comments, votes and survey answers. All of it belongs to your organisation’s project and is visible only to the people that project’s administrators have given access to.
  • Web clippings — The Web Clipper (the browser extension, and the share option in the phone and Mac apps) reads a web page only when you clip it: when you click its button, press its keyboard shortcut, choose “Clip page to Knowledge Commons” or share a page to Knowledge Commons. It sends that page’s address, its text, images and tables, and its citation details (title, author, site, publication date) to your account, with any tags and projects you pick. It does not read your other tabs and it does not record your browsing history. To keep you signed in, the extension stores sign-in tokens in your browser’s extension storage, sends them only to the Knowledge Commons server you chose, and deletes them when you sign out or remove it. A clipping stays in your private Personal Space unless you file a copy into a project, where that project’s members can see it.
  • Connected accounts — If you connect an account you hold with another service (Plaud, Granola, Otter or Zoom), we store the sign-in token that service gives us, encrypted, so we can fetch your recordings when you ask. We never see your password for that service. When you import a recording into a project, we bring in its transcript and speaker labels. If you choose to reprocess it, we also bring in its audio. From then on it is handled like any other recording you upload, under this policy. If you switch on “Send new recordings to my inbox”, we check the account every 15 minutes and keep each new recording’s title, start time and length in your inbox. Nothing else is fetched until you file the recording into a project. Recordings may include the voices of other people, and the same consent responsibility applies as for any recording.
  • Google Drive — If you connect Google Drive, Google lets us reach only two kinds of file: the ones you pick in Google’s own file picker, and the ones Knowledge Commons creates for you. We can’t see or list anything else in your Drive. As with other connected accounts, we store the sign-in token, encrypted. A file you import becomes a document in the project and is handled like any other upload, under this policy. Google Docs, Sheets and Slides come in as Word, Excel and PowerPoint files. When you choose “Open in Google Docs” on a note, wiki page or report, we send that content to Google, which creates a new Google Doc in your Drive. That copy is then yours in your Google account, under Google’s privacy policy, and deleting the original in Knowledge Commons doesn’t delete it.
  • AI assistants — You can connect an AI assistant you use, such as Claude, ChatGPT, Microsoft Copilot or Google Gemini, so that it can search and read your projects. It can never change them. An assistant can reach a project only if that project’s owner has switched on “Allow AI assistants”, which is off by default. Even then, it reaches only the projects you choose when you approve it. Your organisation’s administrators can turn AI assistants off for the whole organisation. Each time an assistant reads something, it goes into the project’s audit log, for example “Claude: searched ‘budget’”. In projects run under the Chatham House Rule, assistants never see speakers’ names.
  • Device identifier — The companion apps generate a random per-install identifier (something like iphone482913) and put it in the name of every audio chunk they upload, so the pipeline can tell one recorder from another. It is not the advertising identifier, it is not shared with anyone, and it is reset if you reinstall the app.
  • Billing information — If you pay for a plan, payment is handled by Stripe (on the web) or Apple (in-app purchase on iOS). We never see or store card numbers. We keep the subscription status, the plan, and a customer or transaction reference so we can tell what your account is entitled to.
  • Technical logs — Standard server logs — IP address, request path, timestamp, user agent and, for signed-in requests, your account identifier — kept for 30 days for security and to diagnose faults, then deleted. We do not use them to build profiles.

How we use it

To run the service you signed up for. That’s the list.

  • Providing the platform — Transcribing recordings, extracting entities and themes, building the knowledge graph, answering questions against it, and showing all of that to the people in your project.
  • Operating your account — Authentication, access control, invitations, plan entitlements and the transactional email that goes with them.
  • Security and reliability — Detecting abuse, investigating faults, keeping backups so a failure does not lose your organisation’s record.
  • Legal obligations — Responding to lawful requests from authorities. We will tell you when we are permitted to.

There is no secondary use of customer data. We do not use your recordings, transcripts or documents to train models, we do not share them with any model provider, and we do not sell them. The one exception is your own choice: when you connect an AI assistant, what it reads from your projects goes to the company that runs that assistant, as described under “AI assistants”. The same goes for anything you open in Google Docs, which goes to Google, as described under “Google Drive”. All inference — transcription, speaker separation, extraction, question answering — runs on GPUs Goal17 operates in Canada, and that processing is stateless: nothing is retained on the inference hardware once a job completes.

Where it lives

In Canada, on infrastructure we operate.

The hosted service runs on servers Goal17 owns and operates in Canada — not a public cloud. Your data is encrypted in transit (TLS). Stored data is protected by physical security and access controls, and we are adding encryption at rest to our storage disks. Backups are kept on Goal17’s own servers in Canada.

A small number of service providers handle specific jobs on our behalf, and each sees only what its job requires:

  • Stripe — Card payments for web subscriptions. Sees your name, email and payment details — which we never receive.
  • Apple — In-app purchases on iOS, handled entirely by Apple under Apple’s privacy policy.
  • Transactional email service — Delivers invitations, verification and billing emails. Sees your email address and the message.

We do not sell personal information, and we do not share it with anyone else except as required by law.

Services you connect to your account are not our service providers. They are where your own recordings come from. We fetch from them only when you ask, or when you switch on inbox sync, and we send them nothing about your projects. The one exception is Google Drive: when you choose “Open in Google Docs”, the note, wiki page or report you chose is sent to Google to create the Doc in your Drive. Your use of each of them is governed by its own privacy policy.

Recordings from Plaud don’t come to us directly from your Plaud account. Plaud only lets other services reach your recordings through its own connection server, so each recording passes through that server, which is in the United States, on its way to us. Plaud states that it doesn’t keep data that passes through this server.

When you connect an AI assistant, what it reads from your projects leaves Knowledge Commons and goes to the company that runs the assistant, such as Anthropic, OpenAI, Microsoft or Google. Where that company processes and stores it, and for how long, is set by its own terms and your settings with it. Knowledge Commons does no AI work on an assistant’s behalf: it only returns what you already have access to.

Your organisation’s role

Your project’s administrators decide who sees what.

Knowledge Commons is a tool organisations use to keep a record of their own conversations. The organisation that owns a project controls its membership, its speaker-identity mode, its retention, and what happens to the content. If you were recorded in a workshop run by an organisation using Knowledge Commons, that organisation is your first point of contact for questions about the recording; we will help them respond.

Your rights

Access, correction, export and deletion.

  • Access and correction — You can see and edit your account details in the app. For anything else, email us and we will send you what we hold about you.
  • Export — A project’s owner can export the whole project at any time from its overview: its recordings, transcripts, documents, notes and data tables, in the same folders as the app. There is nothing to be trapped by.
  • Disconnecting — Disconnect a connected account from your Account page at any time. We ask the service to cancel our access where it supports that, delete the sign-in token straight away, and stop syncing. Recordings you already imported stay in their projects, like any upload, until the project’s owner deletes them. We keep a note of which recordings were imported, and the titles, times and lengths in your inbox, so that reconnecting doesn’t download anything twice. Dismiss inbox entries to remove them. Disconnecting Google Drive doesn’t remove Docs you already opened in Google Docs: they are in your Drive, so delete them there.
  • AI assistants — See the assistants you’ve connected on your Account page, and disconnect any of them there. This cancels its access straight away. A project’s owner can also switch AI assistants off for the project, which cuts every assistant off at once. Anything an assistant has already read stays with the company that runs it, under that company’s terms.
  • Deletion — Delete your account from your Account page and we delete your account information. Content you contributed to a project stays with that project — it is the organisation’s record — unless the project’s owner deletes it. When a project is deleted, it leaves the app at once, and its recordings, transcripts, documents and the graph derived from them are removed from our systems, usually within minutes and always within 30 days. Deleted content ages out of our backups within 13 months.
  • Complaint — If you are not satisfied with our answer, you can complain to the Office of the Privacy Commissioner of Canada, or to your local data-protection authority.

Retention

As long as your organisation keeps the project.

Project content is kept for as long as the project exists, because a permanent record is the point. Account information is kept while your account is active. Billing records are kept for seven years, as Canadian tax law requires. Technical logs are kept for 30 days. Backups are kept for up to 13 months.

Children

Not for people under 16.

Knowledge Commons is a tool for organisations. It is not directed at children, and we do not knowingly collect personal information from anyone under 16. If you believe we have, email us and we will delete it.

Changes to this policy

We’ll say what changed and when.

When this policy changes we update the effective date at the top and, for anything material, email account holders before it takes effect. Previous versions are available on request.

Questions: aaron@goal17.co