1. Scope and responsibility
This Privacy Policy explains how ProInstincts Enterprises, Inc., the operator of Phasoric, handles personal information for Phasoric websites, hosted accounts, and optional connected features. For ordinary direct accounts, Phasoric determines the purposes and means of the account and service data described here. When a customer organization directs us to process personal data in its workspace, that customer is the controller or equivalent decision-maker and Phasoric acts as its processor or service provider under the Data Processing Addendum.
2. Local-first storage boundaries
Storage is selected per vault and applies to routed artifact payloads such as Markdown, attachments, captures, generated media, conversations, transcripts, canvases, governed projections, indexes, and future artifact types.
- Local-only or unmapped vaults: Payloads stay in the browser storage for that Phasoric origin. Creating an account does not upload them or change storage mode.
- Customer-selected providers: Content mapped to Dropbox, Google Drive, or GitHub remains in the authorized device cache and that provider unless you invoke another hosted feature.
- Phasoric Hosted: Content explicitly mapped to Phasoric Hosted is stored in private hosted storage scoped to the account and vault mapping.
- Coordination data: Signed-in workspaces may synchronize bounded metadata and encrypted recovery sidecars so authorized devices can discover, reconcile, and restore a workspace without silently hosting the underlying local-vault payloads.
3. Information we process
- Account and policy data: Name, email address, password hash, verification state, plan and entitlement data, communication preferences, account status, and the policy versions and time accepted.
- Workspace and content data: Content you deliberately host or send to a hosted feature; note and folder metadata such as titles, paths, bounded previews, organizational fields, links, task counts, and timestamps; collaboration, exchange, automation, and connected-service records.
- Connection data: Provider identifiers, scopes, status, encrypted credentials or tokens, synchronization state, and audit events needed for connections you authorize. Secrets are not included in account exports.
- Payment data: Customer, subscription, invoice, transaction, tax, and entitlement identifiers received from Stripe. Phasoric does not receive or store full payment-card or bank-account numbers.
- Signing and exchange evidence: Participant name and email, authentication method, envelope or exchange state, event timestamps, user-agent information, and a one-way network-address hash where used to establish an evidence trail. Encrypted exchange packages and signing documents are stored only for the workflow and retention period described to participants.
- Education and delegated identity data: Organization, roster, course, role, pseudonymous learner, standards, assignment, evidence, review, and integration records authorized by an institution. The Education Data Addendum supplies additional limits.
- Support and feedback: Messages, attachments, contact details, diagnostic information you choose to provide, and an audit record of authorized support access.
- Service and device data: Session and security-cookie state, approximate request and device information, rate-limit and abuse-prevention records, reliability codes, consent state, and bounded first-party signup, billing, activation, and service events. Essential operational events exclude note content, titles, vault names, filenames, searches, private share identifiers, and payment credentials.
- Basic traffic measurement: Aggregate visitor and pageview counts operate independently of optional measurement choices through an analytics infrastructure service. Google Analytics and Microsoft Advertising also receive limited consent-mode signals before optional consent, with analytics and advertising cookie storage denied. Public campaign labels may be included; private link identifiers, arbitrary query values, and workspace content are excluded.
- Optional measurement: With your consent, Google Analytics and Microsoft Advertising may use measurement cookies and receive deeper product or advertising attribution events. OpenAI Ads conversion delivery requires optional consent. Measurement excludes note content, titles, searches, filenames, email addresses, and private link identifiers. You can withdraw optional consent in Settings; baseline traffic counts and limited consent-mode signals continue.
4. Why we process it
We process information to provide requested features; authenticate users; synchronize and recover authorized workspaces; deliver support; secure and troubleshoot the service; administer subscriptions and taxes; communicate operational notices; comply with law; and improve service reliability. Where applicable, our legal bases include performing a contract, legitimate interests in operating and securing the service, consent for optional measurement, and compliance with legal obligations. A customer determines the purpose and legal basis for personal data it instructs us to process.
We do not sell personal information, use private workspace content for personalized advertising, or train a general-purpose model on private workspace content. Source content is treated as user data, not as an instruction that can silently override product controls.
5. Sharing and service providers
We disclose information only as needed to provide a directed feature, comply with law, protect rights and safety, complete a business transaction subject to appropriate safeguards, or with your consent. Categories include hosting and content delivery, managed database and private object storage, transactional email, customer support, payment processing, security and reliability, optional measurement, and hosted-intelligence routing or model processing.
Stripe processes payments. Google Analytics and Microsoft Advertising process the limited consent-mode signals described above and fuller measurement when you grant optional consent. OpenAI Ads receives consented conversion reporting. Other Google and Microsoft integrations process the data needed for features you enable. Dropbox, Google Drive, and GitHub process content you direct to your own connected account under their terms. The public Service Provider & Subprocessor Register describes categories and the process for obtaining the exact current provider list when legally or contractually required.
6. Hosted intelligence and connected providers
Local search, supported transcription, text recognition, and supported local assistance stay on the device. If you enable a hosted route and initiate a request, Phasoric sends the bounded instructions and authorized context needed for that request to hosted-intelligence routing and model services. The route and provider may process that request under their applicable terms. If you connect your own provider, its privacy and retention terms apply. Connected credentials are encrypted at rest, remain server-side, can be revoked in Settings, and are deleted with the account.
7. Cookies and similar storage
Phasoric uses essential browser storage and cookies for sessions, security, preferences, consent, referral attribution, and time-limited protected-share access. Aggregate web traffic measurement does not require analytics cookies. Google Analytics and Microsoft Advertising tags load in a limited consent mode with optional storage denied until consent. Withdrawing optional consent denies that storage, stops deeper product measurement and OpenAI Ads conversion delivery, and clears optional cookies accessible to Phasoric. Basic counts and limited consent-mode signals continue. Withdrawal does not remove data already held by another provider under its retention rules.
8. Retention and deletion
We keep each category only as long as needed for the purposes above, a customer instruction, or law.
- Local data remains until it is deleted on each device, the vault is removed, or browser site data is cleared.
- Hosted account, policy-acceptance, workspace, and Phasoric Hosted content records are deleted through the account-deletion workflow, subject to completion of durable storage cleanup and legally required records.
- Expired verification and reset tokens are cleaned up within 7 days. Webhook events and inbound email operational records are generally purged after 90 days. Transactional email records are generally deleted after 180 days. System alerts are generally kept for 90 days. Backups are retained for no more than 30 days.
- Signing, exchange, billing, security, suppression, and legal records may follow a longer disclosed or legally required period. Where retention is controlled by an organization, requests should be directed to that organization first.
Permanent account deletion immediately revokes sessions, developer and capture tokens, public shares, collaboration access, and active billing. It does not delete local-only content from your devices or content independently stored by Dropbox, Google Drive, GitHub, another participant, or an organization authorized to retain it.
9. Security and international processing
We use access controls, encryption in transit, encrypted secret storage, tenant scoping, audit records, least-privilege processes, security testing, and deletion controls appropriate to the service. No method is perfectly secure. Data may be processed in the United States or another country used by a service category. Where law requires a transfer mechanism, the parties will use an applicable contractual or statutory mechanism; confidential provider and transfer details are available through the DPA process.
10. Your choices and rights
Hosted users can manage connections and optional consent, export hosted account data, separately export local or provider-hosted vaults, and request account deletion in Settings. Depending on applicable law, you may also request access, correction, portability, restriction, objection, or deletion, or complain to a data-protection authority. We may verify identity and may direct an organization-managed request to the organization that controls the workspace.
11. Education and younger users
Phasoric does not invite people who cannot form a direct service contract to create independent hosted accounts. Institution-authorized education use is governed by the institution's instructions and the Education Data Addendum. We do not use covered learner data for personalized advertising or general-purpose model training.
12. Web Clipper
The Phasoric Web Clipper accesses an active tab only when you open it or invoke a capture, highlighting, reader, or OCR command. Depending on the feature, it may process the page title, URL, selected passage, article content, visible-page image, supported PDF content, image URLs, publication metadata, Schema.org data, highlights, and formatting needed to prepare portable Markdown.
Article extraction, template rendering, PDF inspection, and OCR run locally with packaged code and models. Reviewed clips remain in extension storage until you choose to synchronize them to a paired Phasoric workspace. A browser-local vault mirror and agent proposals remain protected by an extension-generated token and explicit approval controls. You can delete clips, reject proposals, rotate the token, export settings, or remove extension-local data by uninstalling the extension or clearing its storage. Information received from Google APIs is handled under the applicable Google API and Chrome Web Store limited-use requirements.
13. Changes and contact
Material changes will be posted with a new version and effective date and, when reasonably practicable, notified through the service or account email. Contact privacy@phasoric.com for privacy questions or rights requests.
14. Connected AI applications, including ChatGPT
When you connect an AI application to Phasoric, it sends the selected tool name and the arguments needed for your request, such as a search phrase, an authorized vault mapping, a note path, a context scope, or content you ask to save. Phasoric does not request your full conversation history. Sign-in and authorization happen on Phasoric; do not paste passwords, API keys, authentication tokens, payment-card data, government identifiers, or protected health information into connected tool requests or notes selected for this integration.
Phasoric checks your account, current plan, granted permissions, and authorized mappings before accessing knowledge. Only explicitly authorized internet-reachable Phasoric Hosted or connected Dropbox, Google Drive, and GitHub mappings are available. Local-only vaults are not remotely exposed. Connected storage providers process the relevant reads or writes under their own terms; their credentials remain managed by Phasoric and are not returned to the AI application.
Depending on the tool you request, Phasoric returns relevant note content, paths, revisions, related knowledge, decision evidence, change history, or a proposed or completed change result to the connected AI application. This information leaves Phasoric for that application. OpenAI handles information returned to ChatGPT under the terms and privacy settings applicable to your OpenAI account. Authorizing a connection does not make your vault public.
Note mutation tools preview by default where supported. Applying a change requires the existing write permissions and transaction safeguards; a client's approval interface does not replace Phasoric authorization. Forge proposals remain subject to their own review controls. Saved notes, proposals, revisions, recoverable deletions and audit evidence follow the retention rules of the relevant Phasoric or connected-storage feature.
Phasoric stores hashed connection credentials, grant/scope information, expiry and last-use times, security audit events and service-allowance accounting needed to operate and protect the connection. The MCP activity record contains a rolling 30 days of daily operation counts and first/last activity, without tool arguments, note bodies, paths or AI responses. Compact account activity remains until account deletion. OAuth security audit events become eligible for bounded cleanup after 90 days; account deletion removes their account association, and scheduled cleanup removes the remaining events. General infrastructure may retain narrowly scoped request/error metadata under the operational and security purposes in this policy; Phasoric does not promise zero retention.
OAuth access tokens last one hour. Continued-access refresh tokens expire after 30 days and rotate when used. Routine bounded cleanup removes expired authorization requests/codes after one day, expired access tokens after seven days, and revoked grants after 30 days. Consumed refresh-token records remain while needed to detect replay against a live successor, then become eligible for cleanup. These are cleanup thresholds; a backlog can delay deletion. Credentials themselves are not included in product analytics or tool results.
To stop a connection, open Settings → AI Connections & API → Connected applications and disconnect its grant. This revokes that grant's access and refresh tokens. Removing a plugin in the AI application should also be followed by this Phasoric-side revocation if you want to verify access has ended; merely removing client configuration is not proof of server-side revocation. Revocation prevents new access but does not delete information already returned to the AI application. Account deletion and export are available in Settings; deletion does not erase independent copies in connected providers or an AI conversation. Contact support@phasoric.com for connection help or privacy@phasoric.com for privacy and data-rights requests.
15. Android and iPhone applications
This policy also applies to the Phasoric Android and iPhone (iOS) applications. Native applications use the account, storage, connected-processing, retention and deletion practices described above. Device features and availability can differ by platform; a native application does not make hosted features local-only.
On Android, text and links you explicitly share to Phasoric are held in an inbox in the application's private device storage until you add or discard them. Receiving a share does not by itself upload a local-only vault or make shared content public. Adding a capture uses the destination vault's existing storage and processing choices. You can remove pending shares in the app; removing app data or uninstalling can also remove device-only data. Keep an export or backup before clearing application storage.
Camera, image, document and audio capture operate when you select the relevant action and grant any required permission. Selected material may then be stored or processed through the local, hosted or connected route you choose. Local reminders are scheduled on your device after the relevant notification permission; their text is not sent to an advertising service. Notification previews may be visible on a device's lock screen according to your Android or iOS settings.
Native store applications do not enable the website's Google Analytics, Microsoft Advertising UET or OpenAI advertising-attribution integrations. Basic aggregate traffic measurement, service reliability and first-party operational events described above can still operate. Optional detailed product-usage reporting follows the existing optional measurement choice. Android install attribution uses a coarse source such as Google Play or other; Phasoric does not request the Android advertising identifier for this purpose. Native activation, capture and retention measurements use bounded event fields rather than raw shared text, note content, URLs or filenames.
When Google Play purchases are available, Google's payment interface handles the payment method. Phasoric receives and processes purchase tokens, product and order identifiers, an obfuscated identifier binding the purchase to your Phasoric account, subscription status and expiry, and the resulting entitlement or credit records. We use these records to verify purchases, restore access, prevent duplicate or fraudulent grants, process purchase lifecycle changes, and provide support. Purchase verification tokens are protected in server storage. Phasoric does not receive your complete payment-card details from this flow. Google processes payment and platform information under its own policies. Necessary financial, security and legal records follow the retention rules above. Deleting the application or device-local data does not cancel a Google Play subscription; its cancellation and account-deletion handling follow the controls and explanations shown in the application.
An AI safety report sends the message and contact information you choose to submit through the in-app feedback service. Your private notes and complete AI conversation are not attached automatically; optional diagnostic summaries remain under your control. Reports are handled for support, safety and service improvement under the purposes and retention practices above. Workspace user-block choices retain the involved account identifiers and workspace relationship to prevent unwanted directed interactions; you can remove a block in the collaboration controls.
To request hosted-account deletion, open Settings → Account → Delete hosted account, or visit https://www.phasoric.com/account-deletion from a browser. The web resource also explains how to request help through privacy@phasoric.com without reinstalling the app. Identity verification, durable deletion processing, retained legal/security records, device-local content and independent third-party copies follow the existing rules in this policy.