PRIVACY · 2026-10-21
Your reading. Your choice.
What we store, why we store it, and how you can change your choices.
Allow this site to record page views, reading progress, a broad referral source and the ID of a prepared editorial post link? Reading and following cards work without this. Hosting-provider traffic logs are separate. Details & settings
Current choice: not selected
Who operates this site
Bream Signals is published and operated by Bream Research Group. Use the contact below for general inquiries and privacy requests.
Contact: contact@dontokois.net
Operator and personal data controller: Hiroyasu Watanabe (trading as Bream Research Group, Japan).
Bream Research Group is the operating name of a sole proprietorship run by one individual in Japan. The proprietor’s address is available without delay on request to the contact above.
Requests and complaints
Requests concerning purposes of use, access, correction, restriction or deletion of personal data, and privacy complaints, can be sent to contact@dontokois.net. We verify identity only to the extent needed for a personal-data request and respond in accordance with applicable law. No identity check is required simply to ask for the proprietor’s legal name and address. Please do not send passwords or medical records.
Contact details and inquiry contents are used to answer, perform necessary identity checks, and handle requests and complaints. We retain them for handling the matter and any applicable legal obligations; you may ask about the retention applicable to your inquiry. We do not sell personal reader data or provide it to advertisers. There is no automated decision-making with legal or similarly significant effects.
Where GDPR or comparable rules apply, optional application analytics rely on consent. Account-linked follows are processed to provide the feature you request; inquiry handling and security serve our legitimate interests in responding and operating the site, subject to applicable law. You may withdraw consent, request access, rectification, erasure, restriction or portability where applicable, object to processing based on legitimate interests, and complain to your competent data-protection authority. Withdrawing consent does not affect earlier lawful processing. You can read without supplying follow or analytics data.
Service providers and location
Hosting and sign-in use OpenAI Sites. The application receives a verified Site user ID and email address when you sign in, and the platform may provide a display name. The email is used for editor access checks, not readership analysis. Follow records use the Site user ID. Sites uses Cloudflare infrastructure. Published Sites data is processed by OpenAI on the operator’s behalf under the Sites Terms and the applicable Data Processing Addendum. The standard Sites DPA also covers Plus and Pro hosting; it is separate from data put into a ChatGPT editing conversation. OpenAI’s published sub-processor list identifies Cloudflare web hosting in the United States and content delivery in locations close to visitors. We do not promise Japan-only processing. The DPA sets out EEA/Swiss transfer safeguards through SCCs or adequacy decisions and UK safeguards through SCCs with the UK Addendum. Provider-side processing is described in the provider’s documentation; this notice does not claim that provider traffic measurement is covered by our optional-analytics switch. OpenAI Sites ↗ · Sites data processing agreement ↗ · Sub-processors and locations ↗
What this site stores
- Follows: the Site-specific signed-in user ID, selected card IDs, explicitly acknowledged versions and follow dates. This supports saved follows across devices; it is not permission for analytics, advertising or email.
- Optional analytics: only after permission, a random browser token, article/card identifier, view or 75% progress event, language, broad source (direct, X, LinkedIn, note, other), and server time. The token is hashed in storage and is not joined to your login account.
- We do not write IP addresses, full referring URLs, arbitrary query strings, device fingerprints, names or email addresses to the application analytics tables. A pseudonymous token is not fully anonymous. We do not infer a reader's health status.
- Daily totals of views and follow/unfollow operations contain no visitor or account IDs. The audience dashboard is restricted to the editor.
Post-specific links
With renewed permission under analytics notice version 2026-10-03.1, we also record a fixed editorial post ID when you land on the matching article from an approved X or LinkedIn link. We do not save arbitrary tracking parameters or carry the post ID to other articles. These counts do not identify people or establish which post caused a follow. Earlier activity is not reconstructed.
Exclude this browser from measurement
For your own checks and editing. A non-unique on/off cookie lasts up to 180 days in this browser only. This does not grant analytics permission or remove earlier totals. Signed-in editor views are already excluded; saved follows remain separate.
Consent, cookies and retention
Optional measurement starts off. Refusal is as accessible as acceptance. Changing to refusal stops new measurement and deletes browser-linked raw analytics. No earlier events are backfilled after later consent. GPC/DNT signals disable application analytics.
The first-party bream_privacy cookie records a random key for your choice for up to 180 days (HttpOnly, SameSite=Lax and Secure on HTTPS). It is only created after a choice. Language is carried in the URL; it needs no preference cookie. A necessary local-storage flag preserves a pending withdrawal if the network fails and is cleared after the server confirms it. A separate local revocation marker coordinates tabs and remains until site storage is cleared; neither marker is sent to the server. Browsers without reliable local storage and cross-tab write coordination keep optional measurement off. Login cookies are handled by the hosting platform.
Raw analytics are eligible for deletion after 30 days; receipts expire after 180 days. Identifier-free daily totals become eligible for deletion after 24 months. Cleanup runs on analytics/privacy/admin API activity; there is no independently scheduled purge, so physical deletion during inactivity is not guaranteed. Follows remain until removed.
Your data
Withdrawal deletes browser-linked raw analytics; identifier-free daily totals remain. Follow data is separate.
Comments and likes · 20 October 2026
Information and purpose
We store a Site-specific account ID, chosen display name, comment text, article and edition, replies, times, review status, rules acceptance and optional AI-use choice. Likes contain the ID, article and time. Reports contain the reporter ID, target comment, reason and time. Limited rate-limit, restriction and moderation records support safety. Ordinary discussion actions use only the Site-specific ID and do not read account emails. Sign-in emails, supplied real names and passwords are not stored in these discussion tables. Approved names, comments and dates are public; account IDs, reports and lists of people who liked an article are not.
Choices and legal basis
Reading requires no discussion account. Posting requires sign-in, a chosen name, text and acceptance of the rules. Participation is for adults aged 18 or older; confirmation is requested, not a birth date or identity document. Where GDPR applies, requested features rely on performance of the service; proportionate abuse prevention and human moderation rely on legitimate interests in protecting participants and the site, subject to your right to object. Optional AI editorial use relies on separate, withdrawable consent. Refusal does not prevent posting. These choices are separate from analytics and advertising.
Who can use the information
The operator reviews submissions and reports. Approved comments are public and may be copied by readers. Only approved comments with editorial-use permission enter the editor-only research export. It excludes account IDs and names, but free text is not necessarily anonymous. This release neither sends comments automatically to external AI nor trains models. Safety review and optional use in new editorial insights are distinct purposes. Before external AI is used for either purpose, its provider and handling must be explained, safeguards and permissions confirmed, and unnecessary personal data removed. We do not sell profiles or a raw comment database. Hosting and sign-in providers are described above.
Retention and deletion
Approved comments remain part of the discussion until removed; necessity is reviewed during editorial updates. Pending, rejected or hidden text is eligible for clearing 90 days after its last update; reports after 90 days; moderation audit records after one year; rate-limit records after 24 hours. Cleanup runs on discussion API activity, not a separate schedule, so inactivity can delay deletion. Likes remain until undone or erased. Restrictions do not expire automatically: necessity is reviewed at least annually and on appeal. Comment deletion clears text, name and AI permission, preserving a minimal thread marker. Account erasure also removes author association, likes and the user's reports; necessary restriction IDs and reasons, and short-lived rate-limit records, can remain.
Copies, hosting and your rights
Withdrawal excludes a contribution from new research exports. Before reuse, the operator must check current permission and remove obsolete working copies. Reader copies and already-published independent analysis cannot be recalled automatically; contact us for review. Provider copies follow the applicable DPA, including return/deletion obligations and legal-retention exceptions; this notice does not promise an unverified backup-purge deadline. We keep requests for provider assistance separate from application deletion. The article panel shows whether submissions are open. Existing follows, analytics choices and platform logs are separate. The controls below export or erase your discussion data; contact us for correction, restriction, objections or other requests. Where GDPR applies, we respond within one month and explain any permitted extension within that period. You may complain to the competent data-protection authority.
Your comments and likes
These controls apply to the signed-in account, not optional reading analytics or saved topic follows.
Hosting is a separate layer
OpenAI Sites hosts the application on Cloudflare-backed infrastructure. Sites automatically provides traffic analytics. This site's consent controls do not claim to stop the platform's security logs, login processing or automatic traffic measurement. Hosting and sign-in processing are also subject to the platform’s terms and privacy information. OpenAI Sites ↗
Links, advertising and notifications
X, LinkedIn and source links open external services only when followed. No embedded social feeds, advertising SDKs or marketing pixels have been added. Saving a card does not subscribe you to email or push notifications.