Privacy
The short version
We hold what we need to scan your repositories and show you the results: who you are, which repositories you connected, the credentials you gave us for those repositories, and the reports we produced.
Your code is not intentionally executed. It is read and scanned. A scan on its own never reaches an AI provider. Only when you ask for an AI Review, in the app or from an AI tool you connected, do that scan's findings go to Anthropic, each with the code around where it was found (up to 80 lines), never your whole repository.
Who we are
Tuvio is a product of CodexMotive Digital Solutions, a sole proprietorship registered in Ontario, Canada, based in Waterloo, Ontario.
For anything about this policy or your personal information, contact info@tuvio.dev.
Information we collect
- Account and identity information. You sign in with GitHub or Google and we receive an account identifier, a display name and, where the provider gives it, an email address. GitHub accounts may withhold the email; the service works without it. We use this to know who you are and to secure your account.
- Free AI review fingerprint. We no longer create this for new accounts. For accounts that had a free AI review, it is a one-way fingerprint of the ID of the GitHub or Google account you sign in with, which stops that review being claimed again. It is a keyed hash: it contains no email, name or other detail, and it can only be matched against a sign-in ID by us, with a secret key we keep. Ones made earlier are kept until 2 years after the account is deleted, then deleted, and we use them for nothing else.
- Network fingerprint and GitHub account age. To stop one network from using many free scans or free AI reviews, we keep a keyed one-way fingerprint of the network address a scan or an AI review request came from, never the address itself, and the date your GitHub account was created. A scan's fingerprint is deleted as soon as the scan finishes; an AI review's is kept for 2 days. Every fingerprint is deleted with the project or your account. Scans run without an account work differently, as the next item says.
- Scans of a repository without an account. When you scan a repository without signing in, we set a cookie holding a random ID (see "Cookies and analytics") and store, with the scan, the repository address, the branch, the report, a one-way hash of that cookie's ID, and a one-way SHA-256 hash of your IP address combined with that cookie's ID. The first hash lets that browser see the scan's full report and claim the scan when you sign up; we clear it within 32 days, since the cookie lasts 30. We use the second to count how many distinct browsers ran such scans in the last 48 hours, a usage statistic we compute whether or not you gave analytics consent, and we clear it within 3 days. If you give us a token to reach a private repository, it is stored encrypted only until the scan ends, and we keep a one-way hash of it after that. To limit free scanning, we also count these scans per IP address and per browser, in the server's memory, for up to a day. A scan can be claimed by signing up for 89 days; one nobody claims is deleted, with its stored files, within 91 days. The scan request also runs Google reCAPTCHA (next item).
- Google reCAPTCHA. When you start a scan without an account, the page loads Google reCAPTCHA v3, which collects information about your browser, device and how you interact with the page, and gives the request a score from which we decide whether it looks automated. Google processes this under its own Privacy Policy and Terms.
- Repository information — the URL, the provider, and the branch you asked us to scan. We need this to know what to scan and where.
- Repository credentials that you supply, encrypted at rest, used only to contact your repository provider on your behalf to run the scans you request.
- Scan data and reports — findings, the files and lines they point at, and, once a scan has been AI-Reviewed, the fix prompts it generated. Every scan also stores the code around its top-ranked findings so that an AI Review can be run on it later; see "Repository and source-code processing".
- AI Review data. When you ask for an AI Review, a completed scan's findings, each with the code around where it was found, are sent to our AI provider. See "AI Review" below.
- Monitoring, if you turn it on for a repository. We scan it again every week and keep a record of the findings we've already seen (the kind of issue and the file it is in), so that we email you only about new High or Critical issues.
- Connected AI tools, if you connect one (for example an AI coding assistant, through our MCP server). We keep the name you give the connection, a one-way hash of its token (never the token itself), the token's last four characters, when it was created, last used and expires, whether you let it spend AI Reviews, and your monthly limit for that. What the tool fetches from us, such as findings and fix guidance for your repositories, goes to that tool and its provider and is handled under their terms, not ours.
- Support requests. A request you send from the app is emailed to our support inbox with your account's ID and email address; we don't store it in the app. When you sign up, our inbox also receives an email with your account's ID and email address.
- Your email consent, if you opt in to product news. We keep whether you agreed, the email address and the version of the wording you agreed to, where you gave it, and when, and a record of each later change. See "Emails".
- Ratings and testimonials, only if you give them. When you rate an AI Review, we keep your rating and any comment with your account; comments are only read by us. If you send a testimonial, we keep the quote, the name and role you entered, how you chose to be credited, and when you agreed to it being shown. We keep the full name you enter; if you chose "First name and initial", only your first name and last initial are shown. We show a testimonial on our website only after you agreed and we approved it, credited the way you chose.
- Billing information if you buy something. Stripe, our payment provider, holds your payment details; we never see your card number. We keep a record of each purchase and refund (what you bought, the amount and currency, and Stripe's references for the payment), your Stripe customer and subscription IDs, and, where you confirmed at checkout that you want immediate access, which version of that confirmation you accepted and when.
- Analytics and cookie information, only once you grant analytics consent.
- Technical, log and security information — requests, errors, timestamps and IP addresses our infrastructure generates in the ordinary course of running the service.
We do not sell any of it, and we do not use your source code, your findings, or your prompts to train any model of ours.
How we use information
Authentication, running the scans you request, generating reports, maintaining scan history, running AI Review when you ask for one, monitoring the repositories you choose, serving the AI tools you connect, sending the service emails described under "Emails", securing the service against abuse, troubleshooting, billing, and customer support are things we do because they are necessary to provide the service you signed up for. Keeping purchase and refund records is also something tax law requires of us. Google Analytics is different: it only loads once you've given consent, and stops when you withdraw it. Some of our abuse checks make automatic decisions; see "Your privacy rights" for which ones and how to ask a person to review one.
Repository and source-code processing
A scan clones your repository onto a temporary, isolated worker, runs the scanner tools over the files, writes a report, and the worker is destroyed.
Your repository is not intentionally built, installed, or executed. The runner does not install dependencies, and does not intentionally run a project's postinstall script or any other project-defined script — from your repository or anyone else's. This is a deliberate boundary in how the scanner is built, not a promise that no bug in any tool could ever cause unintended execution.
We keep the report. We do not intentionally keep a copy of your repository after the scan completes. Because a finding has to point at something to be useful, a report can include short excerpts or identifiers drawn from your repository's contents. We don't promise a finding can never contain a fragment of source code, only that we don't retain the repository itself.
The code kept for AI Review. So that you can ask for an AI Review later, every scan also stores, in our cloud storage, the code around each of its top-ranked findings: up to 80 lines and 4,000 characters per finding, for up to 150 findings. No code is stored this way for a leaked-key finding. This is the one place we keep blocks of source code from your repository, and it is deleted 30 days after the scan, which is why a scan older than that can't be AI-Reviewed. It never goes to an AI provider unless you ask for a review.
npm package lookups. To check that your project's npm packages exist, our servers look up the names of its direct npm dependencies on the public npm registry (registry.npmjs.org and api.npmjs.org), with the user agent tuvio-repo-scan. Nothing else from your repository is sent. Packages your project installs from a private registry are never looked up.
We record the names of npm packages your project lists that don't exist on npm, without your repository or account, to track package names AI tools invent. A name is only used once it appears in projects of at least three different accounts or organisations. We delete a name's record 12 months after we last saw it.
Live site scans
When you paste a website's address, we read that site's public pages the way a browser does: its home page, the scripts it loads from the same site, robots.txt, and a few well-known paths, with plain GET requests. We don't log in, submit forms or send cookies to the site. What we fetch is listed on how our scanner works.
What we collect for a site scan, and why.
- The address of the site you ask us to scan, and the report we write about it: to run the scan and show you the result.
- A keyed hash of your IP address, never the address itself: to limit free scanning and to protect the service.
- A random ID in a cookie, which we keep only as a hash: to know which scans this browser ran when you have no account.
- If you're signed in, your account, including its email address: to keep your scans in your account. We email you about Tuvio only if you opt in.
The report goes to whoever asked for it, who may not own the site. You may only scan a site you own, are authorized to test, or check only to tell its owner (see the Terms). Site scans never go to Anthropic or any other AI provider.
What we keep. The report: the kinds of problems found and where (an address on the site, a header's name and value, a cookie's name, the kind of key), the site's address, and when it was scanned. We don't store the pages or scripts we read, and we never store or show the value of a key we find.
Who can see it. Only you: the browser that ran the scan, or your account once you sign in. If you make a share link, anyone with that link can see the report until the link expires (at most 7 days, and never after the scan is deleted) or you revoke it. We keep only a one-way hash of the link's token. A finished scan of a site may be reused for 24 hours to answer another request for the same site; that reuses what we found about the site, never who asked.
Scanning without an account. The first scan sets a cookie on api.tuvio.dev that makes the scan yours in that browser. It lasts 30 days and holds a random ID, nothing else. Signing in moves those scans into your account. To limit free scanning, we count scans per IP address, per network and per sign-in, over a rolling 24 hours. For that and for security, we keep a keyed hash of the requester's IP address (never the address itself): the scan record clears it after 48 hours, and security logs keep it for 30 days.
Verifying a site you own. When you verify a site in a project, our scanner sends automated GET requests to a fixed list of paths on that site (for example /.env and /.git/config, well-known API paths such as /openapi.json and /graphql) and to AI endpoints the site's own code names. Before these checks start, we email the site's security contact (the address in its security.txt, otherwise security@ and admin@ at its domain). For each of these requests we read at most the first 2 KB of the response; we keep only the path, its status code and the kind of finding, never the response's content. If the site's own code names a backend service its owner has confirmed, our scanner also counts the rows and files that service exposes to the public, without reading them.
Authorization record. When you verify a site, we keep a record of the proof (how and when control of the site was shown, a fingerprint of the token or DNS value that matched, and the server address that served it), of every scan queued to probe the site or its confirmed backend, of every probe run, and of each backend you confirmed or removed, with the consent wording you agreed to. It holds your account id and the site's address, never your email, your name or what the site returned. It is kept while the site stays verified or a backend of it is confirmed, and for 12 months after its last check; it is kept for that time even if you delete the site, the project or your account.
Site notice records. When we email a site's security contact about checks on that site, we keep the domain, when the notice was queued and how many addresses it went to (not the addresses) for 97 days.
How long. Site scans are deleted 7 days after they run, or 30 days on Solo.
Scans in your projects. We keep the latest scan of each site in your projects while it is there; other site scans follow your plan's retention.
AI Review
What is sent, and when
AI Review sends one completed scan's findings (up to 100) to Anthropic, each with the code around where it was found (up to 80 lines, at most 4,000 characters). It runs when you ask for one, in the app or from a connected AI tool, or when you buy a 14-day pass, which starts its review on the report you bought it from. It uses a review credit you hold, the review a pass includes, or your Solo allowance. A scan on its own never reaches an AI provider. A file that holds a credential we detected is not sent at all. It never sends your whole repository, and any credential is masked before it leaves.
Which provider
AI Review goes to Anthropic. A scan goes nowhere near an AI provider: it runs scanners over your files, and what it keeps stays in our own systems.
What we promise, and what Anthropic's own terms promise
We do not use your source code or scan findings to train any model of ours — that is a statement about what we do. Separately, our contract with Anthropic, PBC is governed by Anthropic's Commercial Terms of Service and its Data Processing Addendum (DPA), which is automatically incorporated into that agreement. Under those terms, as published today:
- Training: Anthropic's Commercial Terms state that Anthropic may not train its models on the content we send it through the API.
- Retention: Anthropic states that it deletes API inputs and outputs within 30 days of receiving or generating them, unless it needs to keep them longer to enforce its Usage Policy (content its automated systems flag can be kept for up to 2 years) or to comply with the law. Its DPA also commits to deleting customer data within 30 days of the agreement ending, subject to legal retention exceptions.
- Processing location: Anthropic states that data is stored in the United States. By default, traffic may also route to the US, Europe, Asia, or Australia for processing, and internal processes such as safety review or support may occur in other countries where Anthropic or its affiliates operate. Anthropic's DPA incorporates Standard Contractual Clauses for EU/UK/Swiss transfers as a safeguard for this.
Anthropic's terms can change; this reflects what's published as of this policy's last update, not a permanent guarantee about a third party we don't control.
Repository credentials
Encrypted at rest. No endpoint of ours returns a stored credential in any form. The interface shows you the last four characters so you can recognise which credential is which, and nothing more.
Disconnecting a repository or deleting a project removes its stored credential from our active systems. Our database backups keep a deleted credential for up to 7 days before it is gone from them too; see "How long we keep information" below.
Service providers
Their role differs by function, and so does the relationship:
- Microsoft Azure — hosting, storage and message queues, as our infrastructure provider. Region: Canada and the United States.
- Anthropic — the AI provider for AI Review. A scan does not reach them, and neither does a live site scan. See "AI Review" above for what governs this relationship.
- Microsoft 365 — the email we send you, such as monitoring alerts and our replies, through Microsoft Graph, and our inbox, which receives support requests and sign-up notices.
- Stripe — payments, if you buy something. We contract with Stripe Payments Canada, Ltd., though Stripe's own processing occurs primarily in the United States, as described under "International transfers" below. Stripe collects your payment details directly; its own privacy policy governs that.
- GitHub and Google — sign-in.
- Your git provider — GitHub, GitLab, Azure DevOps or Bitbucket — which we contact with the credential you supplied.
- Google reCAPTCHA: telling people from automated traffic when you start a scan without an account.
- Google Analytics — usage analytics, only once you grant analytics consent.
- An AI tool you connect: not our service provider. What it fetches from us goes to that tool and its provider at your request, under their terms.
Where a provider processes information on our behalf for our purposes, we treat it as our processor. Where a provider has its own independent relationship with you — Stripe processing your payment, or GitHub/Google as your account provider — it may also act as an independent controller for its own purposes.
International transfers
Your information may be processed outside the province or country where you're located, because our service providers operate infrastructure in multiple locations:
- Azure hosting/storage/worker regions: Canada and the United States.
- Anthropic: data is stored in the United States; traffic may also route to the US, Europe, Asia, or Australia for processing by default.
- Stripe: we contract with Stripe Payments Canada, Ltd., but its own DPA states personal data is transferred to Stripe, LLC in the United States, and may reach its other affiliates and subprocessors in other jurisdictions, under Standard Contractual Clauses or the Data Privacy Framework.
- Google Analytics: processed by Google, a U.S. company with globally distributed infrastructure. International transfers rely on Google's certification under the EU-U.S. and Swiss-U.S. Data Privacy Framework, and on Standard Contractual Clauses where that framework doesn't apply.
- Google reCAPTCHA: processed by Google in the United States and wherever else Google operates, under Google's own terms.
Azure processing occurs in the United States as well as Canada, and Anthropic, Stripe and Google may also process information outside Canada. Information processed in another country is subject to that country's laws, and its courts, law enforcement and national security authorities may be able to access it there.
Security
We use safeguards designed to protect your information, including:
- encryption at rest for repository credentials;
- restricted internal access to stored credentials and repository data;
- ephemeral, isolated infrastructure for running scans, destroyed after each scan completes;
- a scanning design that does not intentionally execute repository code;
- no API endpoint that returns a stored credential in usable form.
No online service can guarantee absolute security. We work to protect your information using reasonable safeguards, but we cannot promise that unauthorized access, loss, or disclosure will never occur.
How long we keep information
| Information | Retention |
|---|---|
| Account information | While your account exists, subject to legal retention requirements |
| Free AI review fingerprint (no longer created for new accounts) | While your account exists, then 2 years after you delete it |
| Network fingerprint of a scan request | Until the scan finishes |
| Network fingerprint of an AI review request | 2 days, and deleted with the project or your account |
| GitHub account creation date | While your account exists |
| Repository connection details | While the repository stays connected |
| Repository credential/token | While connected, subject to backup retention below |
| Scan reports and AI Review results | While the related account/project exists |
| Code kept for AI Review (up to 80 lines per finding) | 30 days after the scan, then deleted |
| Temporary repository clone | Only for the duration of the scan that used it |
| Names of npm packages that don't exist on npm, without your repository or account | 12 months after we last saw the name |
| Scans run without an account | Deleted within 91 days, with their stored files, unless you claim one by signing up within 89 days; a claimed scan becomes part of your account |
| Hash of your IP address and cookie ID on a scan run without an account | Cleared within 3 days |
| Hash of your cookie ID on a scan run without an account | Cleared within 32 days |
| Internal copies of emails and jobs we sent (may include repository addresses) | Deleted within 31 days |
| Anonymous scan counts per IP address and per browser | Up to a day, in the server's memory |
| Monitoring's record of findings already seen | While monitoring is on, then 30 days after you turn it off; deleted with the project or your account |
| Connected AI tool records | Until you delete your account, including after a token expires or you revoke it. A token expires 90 days after it's created (30 days on the Free plan) |
| Support requests and sign-up notices in our inbox | No set expiry today; email us to have yours deleted |
| Your email consent and its history | While your account exists |
| Live site scan reports | 7 days after the scan (30 days on Solo). Without an account, readable only in the browser that ran them |
| Latest scan of each site in your projects | While the site is in the project; other site scans follow the row above |
| Authorization record of a verified site (account id, site address, how and when the site was proved, a fingerprint of the matching token or DNS value, the server address that served it, scans queued to probe, probe runs, backends confirmed or removed and the consent wording) | While the site stays verified or a backend of it is confirmed, and 12 months after its last check; also after you delete the site, the project or your account |
| Site notice records (domain, when the notice was queued, how many addresses it went to) | 97 days |
| Site scan share links | Until they expire (at most 7 days) or are revoked; deleted with your account |
| Site scan counts per IP address, network and sign-in | A rolling 24 hours |
| Site monitoring's record of findings already seen | 90 days after a finding was last seen, while weekly checks of the site go on |
| Keyed hash of a site scan requester's IP address | 48 hours on the scan record; 30 days in security logs |
| Security logs of site scans | 30 days |
| Logs | 30 days |
| Database backups | Up to 7 days (point-in-time restore), stored in Azure Canada Central with a copy in Canada East; no longer-term backups |
| Stored files (scan and site-scan reports, code around findings) | No backups: a deleted file is gone at once |
| Billing records | The app's own copy: while your account exists. The payment and refund records in our Stripe account: as long as Canadian tax law requires us to keep business records, generally 6 years from the end of the tax year they relate to (Canada Revenue Agency rules) |
When you delete a project or account, the corresponding information is removed from our active, production systems, as described below. Information deleted from our database can remain in its backups for up to 7 days, until those backups roll over. Stored files, such as reports and the code saved around findings, have no backups. We don't describe deletion as immediate everywhere, because for the database it isn't.
Deletion
Deleting a project removes its scan history and its stored repository credential from our active systems. Deleting your account removes your projects, scan history, repository credentials, AI Review results, monitoring records, connected AI tools, the app's own billing records, your email consent record, ratings and testimonials from our active systems.
Your site scans are unlinked from you at once and deleted within 7 days (30 for Solo). Their share links stop working at once. Security logs keep their IDs for 30 days. If you verified a site, its authorization record (account id, site address, proof, times) is kept while the site stays verified or a backend of it is confirmed, and for 12 months after its last check, even after your account is deleted.
To have a published testimonial taken down, email info@tuvio.dev. We remove it from the website and stop using it. If you sent a testimonial, the email we received about it (your quote, name and email address) stays in our inbox after you delete your account; email us and we'll delete it.
Some things are deliberately kept after account deletion. The record of each payment and refund stays in our Stripe account for as long as tax law requires (see the table above). The free AI review fingerprint described under "Information we collect", if your account has one, stays for 2 years so that review can't be claimed again with the same sign-in. Logs keep what they recorded for their 30 days, and emails in our inbox, such as support requests and your sign-up notice, stay until you ask us to delete them. Information may also persist briefly in backups, as described above, until those backups cycle out in the ordinary course.
Your privacy rights
We're based in Ontario, which has no general private-sector privacy statute of its own — so the federal Personal Information Protection and Electronic Documents Act (PIPEDA) governs our handling of personal information directly. If you're a Quebec resident, Quebec's Law 25 also applies to you, regardless of where we're based, because it protects Quebec residents based on where they live, not where we operate.
Subject to applicable law and reasonable exceptions, you generally have the right to:
- ask what personal information we hold about you and how it's used;
- correct information that's inaccurate or incomplete;
- withdraw consent for consent-based processing, such as analytics;
- request deletion, as described above;
- if you're a Quebec resident, receive the personal information you gave us in a structured, commonly used technological format, or have it sent to someone you name;
- raise a concern with us directly.
How to ask. Email info@tuvio.dev. We'll confirm it's you before we answer, for example by asking you to write from the email address on your account. We answer within 30 days, at no charge. If the law lets us take longer and we need to, we'll tell you why within those 30 days.
Automated decisions
Some decisions are made automatically, without a person looking at them first:
- Scans without an account: a request is refused if Google reCAPTCHA scores it as likely automated, or if the IP address or browser has reached its scan limit.
- Scans and AI reviews when you're signed in: a request can be refused or delayed when your account, or the network it comes from, has reached a limit on how many scans or free reviews can run.
- The free AI review: whether you can use it yet depends on how long ago your GitHub account was created (or, if you signed in another way, your Tuvio account), how many free reviews have started today from the same network, and how many have started today across the service.
The information used is the request itself, the reCAPTCHA score, the keyed fingerprint of your network address, your account's age and your GitHub account's creation date. If one of these decisions affected you, you can ask us, at info@tuvio.dev, what information was used, the reasons and main factors behind the decision, and to correct that information. You can also send us your comments and ask for the decision to be reviewed by a person who can change it.
If there's a breach
If a breach of our safeguards involving your personal information creates a real risk of significant harm to you, we'll tell you as soon as we can, report it to the Office of the Privacy Commissioner of Canada, and, where the information is about Quebec residents and the risk is of serious injury, to the Commission d'accès à l'information du Québec. We keep a record of every such breach, as the law requires.
Not every right applies to every user in every jurisdiction; which rights apply depends on where you're located and which law governs your relationship with us.
Users outside Canada
This applies only to the extent the relevant law actually reaches you.
EEA / United Kingdom
If GDPR or UK GDPR applies to you, we act as controller of your personal information. Our lawful bases are performance of a contract, legitimate interests (security, fraud prevention, service operation), legal obligation (keeping purchase and refund records for tax law), and consent (analytics cookies). We're based in Canada, which the European Commission and the United Kingdom recognise as giving adequate protection to personal information handled by businesses under PIPEDA; where our providers process it elsewhere, see "International transfers". You have the rights to access, correct, erase, restrict and port your information, to object to processing based on legitimate interests, to withdraw consent, and to lodge a complaint with your local supervisory authority. International transfers rely on appropriate safeguards — see "International transfers" above.
California
The California Consumer Privacy Act only applies to businesses that meet certain revenue or data-volume thresholds. Given our current size, we don't believe we meet them — but if we ever do, California residents will have rights to know, delete and correct their personal information, to opt out of sale or sharing (we do not sell or share personal information), and to non-discrimination for exercising these rights, and this section will be updated to reflect that.
Emails
Service emails go to the email address on your account: monitoring alerts about new High or Critical issues, replies to what you send us, and notices about your account, these terms or this policy. Monitoring alerts have an unsubscribe link, and you can also turn them off under Settings → New-issue emails. Notices that monitoring has paused are still sent, because they ask you to act.
Marketing emails, such as product news, go only to people who have opted in. We don't send any today.
You can opt in when you sign up or under Settings → Product news emails. The box is never ticked for you. We keep a record of your consent: the email address, the version of the wording you agreed to, where you gave it and when, and each later change. Every marketing email has an unsubscribe link, and you can withdraw your consent at any time by turning off Product news emails in Settings.
Children's privacy
The service is not directed to children, and we do not knowingly collect personal information from children under 13 (under 14 in Quebec), or under the higher minimum age set by the law where they live. If you believe a child has provided us with personal information, contact us and we'll take appropriate steps to remove it.
Your repositories
You're responsible for having the right to connect a repository, provide its access credentials, have it scanned, and authorise us to process whatever information it contains — including sending scan findings for AI Review where applicable. The corresponding authorisation and acceptable-use commitments live in our Terms, since they describe what you're agreeing to by using the service. This page describes what happens to information once it's in our systems.
Changes to this policy
We may update this policy as the service changes. If a change is material, we'll notify you before it takes effect — by posting a notice on this page and, where practical, by email or in-product notice. The date at the top always reflects the current version.
Contact and complaints
CodexMotive Digital Solutions, registered in Ontario, Canada, based in Waterloo, Ontario. Privacy enquiries: info@tuvio.dev.
Who is accountable. The business owner is accountable for how we handle personal information and acts as our privacy officer. Write to info@tuvio.dev to ask what we hold about you, to correct it, to withdraw consent, or to complain.
If you raise a concern and aren't satisfied with our response, you may escalate it — in Canada generally, to the Office of the Privacy Commissioner of Canada; if Quebec's Law 25 applies, to the Commission d'accès à l'information du Québec; if you're in the EEA/UK and GDPR/UK GDPR applies, to your local supervisory authority.
Related reading: Terms and the questions page, which answers most of this in plainer language.