Privacy Policy

Last updated: September 7, 2026

This policy covers blury.ai and its subdomains. The operator trading as Blury, based in Ontario, Canada, is responsible for the personal information described here. The operator is also the privacy contact: hello@blury.ai. Where applicable law uses those terms, the operator acts as controller for account and Service administration. A business customer’s separate signed processing agreement may govern content processed on its instructions.

Before submitting writing: the features you request may send your content to OpenAI and/or Weights & Biases (W&B/CoreWeave). Saved writing, Style and Memory can persist in your account and browser. Blury does not use private customer content to train its own models; provider terms differ, as explained below. Do not submit highly sensitive or unauthorized information.

1. Information and sources

Account and authentication: email, display name, optional picture, password hash for password accounts, Google sign-in identifiers/profile information and authorization records, sessions, IP address and browser information associated with authentication, and two-factor secrets and recovery information if enabled. Information comes from you, your browser and your chosen sign-in provider.
Writing and files: prompts, conversations, documents, uploaded files and images, imported text, samples, Style preferences, Memory entries, generated results, detector results and share records. Information comes from you and from the processing you request and may include information about other people.
Billing and usage: Stripe customer, subscription and transaction references, plan and payment status, trial/renewal dates, invoices and limited billing details, words/tokens, request counts, usage totals and timestamps. Stripe receives payment details directly; Blury’s application does not collect full payment-card numbers.

1. Information and sources

Account and authentication: email, display name, optional picture, password hash for password accounts, Google sign-in identifiers/profile information and authorization records, sessions, IP address and browser information associated with authentication, and two-factor secrets and recovery information if enabled. Information comes from you, your browser and your chosen sign-in provider.

Writing and files: prompts, conversations, documents, uploaded files and images, imported text, samples, Style preferences, Memory entries, generated results, detector results and share records. Information comes from you and from the processing you request and may include information about other people.

Billing and usage: Stripe customer, subscription and transaction references, plan and payment status, trial/renewal dates, invoices and limited billing details, words/tokens, request counts, usage totals and timestamps. Stripe receives payment details directly; Blury’s application does not collect full payment-card numbers.

7. Retention and deletion

We retain information according to its purpose, legal requirements and the controls below. A configured cleanup threshold is not always an exact physical-erasure deadline: some cleanup runs on later activity or maintenance. Browser copies and provider records have separate lifetimes.

Active account and saved work: retained while needed for your account or until you delete the relevant item/account. Chat, documents, samples, Style and Memory have separate controls. Deleting one does not necessarily remove independent copies in another. Activity history is bounded to 24 recent saved results and can contain complete text.

Generation recovery: a separate bounded cache of up to 24 recent generation results has a 24-hour lifetime. It is excluded from user-state backups. Saved history or documents may retain the same result longer.

Presence: records become eligible for cleanup 24 hours after last activity; the browser identifier is treated as valid for 24 hours. Browser local storage does not automatically erase itself at that moment.

Generation-error diagnostics: the application removes rows older than 14 days when logging subsequent errors. Those records contain account references and bounded diagnostic fields; that is not a 14-day promise for all infrastructure logs.

Authentication rate limits: cleanup uses a 24-hour threshold, with older records pruned during activity. Other action/usage counters have their own rolling windows and may remain with account state. Authentication sessions and their IP/browser metadata persist until expiry or revocation and subsequent cleanup.

Trial claims: the keyed mailbox identifier, previous account ID and claim timestamp survive deletion. The current system has no automatic expiry for this narrow anti-abuse record. You can ask for review or deletion; any refusal must have a lawful reason. Stripe may retain customer and transaction history independently.

Recovery copies: configured recovery schedules are 31 days from creation for immutable file versions, 35 days for user-state snapshots and 90 days for D1 recovery exports. Copies are restricted and isolated from the live website. They are not selectively edited when live records are erased.

Billing, tax, support and legal records: retained where needed for the transaction, an unresolved request, fraud prevention or a legal obligation. Canadian tax records generally require retention for six years from the end of the relevant tax year, and some records require longer retention. Support mail and legal records do not currently have a universal automatic deletion period. Retention must be limited to a justified purpose.

Account deletion starts a durable, retried workflow. It blocks new sign-ins and checkout, confirms subscription cancellation, revokes sessions, drains already authorized requests, then erases live account data and files. This is not instantaneous. A signed-out status page reports progress; a completion email is attempted after live erasure. A billing or infrastructure failure can delay completion. Contact us if the process stalls.

9. Your choices, requests and complaints

Email hello@blury.ai to request access, correction, a copy, deletion or review of processing. You can also edit content and request account deletion in account settings. Describe the request and relevant account; do not send a password, full card number or identity document unless we explain a necessary secure verification process. We use proportionate verification and may need proof of an authorized agent’s authority. People whose information appears in another user’s content may also contact us.

Rights depend on applicable law and may have exceptions. We explain a refusal or limitation where required, provide any available appeal route, and do not retaliate for a lawful request. Required service processing may become impossible after deletion or withdrawal. Requests are generally free, subject only to lawful exceptions.

Canada: applicable federal/provincial law may provide access, correction, withdrawal of consent and complaint rights. Under PIPEDA an access response is generally due within 30 days, with only permitted extensions and notice. You may complain to the Office of the Privacy Commissioner of Canada or the relevant provincial authority, including Quebec’s Commission d’accès à l’information.

EU/EEA and UK: where GDPR applies, rights include access, correction, erasure, restriction, portability where applicable, withdrawal of consent and objection to legitimate-interest processing. We respond without undue delay and ordinarily within one month; a permitted extension requires notice and reasons. You may complain directly to your supervisory authority or the UK ICO, without contacting us first.

California and other U.S. states: where an applicable state privacy law covers Blury and your request, rights may include knowing/accessing, correcting, deleting or obtaining portable information, opting out of specified processing, using an authorized agent, non-discrimination and appealing a refusal. California’s response period is generally 45 days, with a permitted notified extension. State applicability thresholds and exceptions differ; this policy does not assert that every state law applies to every account.

The categories we collect and disclose for business purposes are identifiers, account/commercial information, internet/network activity, approximate country, user content and related inferences, and any sensitive information users submit contrary to our request. Their sources, purposes, recipients and retention criteria appear above. We do not use sensitive information to infer personal characteristics for advertising, and we do not sell or share information for cross-context behavioural advertising. A Global Privacy Control signal does not need to opt you out of a sale or advertising-sharing practice we do not operate; use the analytics controls for optional presence measurement.

Privacy Policy

1. Information and sources

Privacy Policy

1. Information and sources

2. Purposes and legal bases

3. AI providers, personalization and training

4. Recipients and sharing

5. Optional Google Drive import

6. Cookies and local storage

7. Retention and deletion

8. Processing locations and security

9. Your choices, requests and complaints

10. Children and automated processing

11. Changes and contact

Legal