Policies

Your dreams are private by default.

This policy explains what we collect, why we use it and the choices you have. It covers the current website and beta features operated by Uncanny Post Press, LLC.

Last updated September 23, 2026

On this page

Who handles your information

Uncanny Post Press, LLC operates The Uncanny Post. We use information to provide the features you request, secure accounts, operate the beta and respond to support requests. Contact help.me@uncannypost.com with privacy questions. Ordinary journal storage uses account access controls; it is not end-to-end encrypted. We do not sell private dream content or use it for advertising.

Beta applications and testing

An application records your email, questionnaire answers, device information, interests, participation choices, beta-email consent, agreement version, individual acknowledgments and acceptance time. Invitation administrators review these records. Applying does not create an account or subscribe you to marketing. We use keyed network fingerprints to limit abuse; raw network addresses are not stored in waitlist records. The beta program records assignments, credit, submitted feedback, reports, wishes and votes, team responses, notices and delivery state, pauses, reviews and appeals. Testing feedback is private to you and authorized staff; a wish is shared only when you choose. You may test with labeled sample dreams. To request deletion of a separate application, email help.me@uncannypost.com from the submitted address; deleting an account does not automatically remove an earlier application.

Help and support

Support requests are separate from your journal and visible to you and authorized site administrators. Responses appear on the support page. When email delivery is available, we may send a generic update notice; return to Support to check your request. Account-linked support records are deleted with your account. Email correspondence is separate from the in-site request history. Do not include passwords or unnecessary sensitive information. Support messages are not used for model training, research or marketing. This service is not continuously monitored.

Optional dream profile and nightly check-ins

You can choose to save a private profile of dreaming habits and dated notes about sleep, stress, senses, recent experiences, food or drink, and wellbeing. Every question can be skipped. These answers are stored separately from dream narratives and are not sent to AI, used for marketing, incorporated into The Corpus, or shared for research. They are personal reflection records, not clinical assessments. Turning off the profile and saving erases its answers. Turning off an optional check-in notes section and saving erases those notes. You can delete individual check-ins, download selected categories to share yourself, or include them in your full journal export. Account deletion removes these records. Downloaded copies remain under your control.

Other private reflection records

You may save sleep records and imported CSV rows, life events and chapters, place sketches, inspiration images and links, creative projects and personal experiments. These are private reflection records, not clinical assessments or proof of cause and effect. Current sleep inputs are manual or imported; they do not imply a connected sleep tracker or precise location collection.

Dictation and original recordings

Browser dictation produces text; your browser or its speech-recognition provider may process audio. The Uncanny Post does not save dictation audio. A separate microphone-recording feature lets you record, preview and choose to save or download original audio privately. Saved original recordings are not sent for transcription or AI interpretation. Microphone access requires your browser permission.

How a requested interpretation is processed

The site’s interpretation engine processes the dream and context you submit, together with relevant dictionary material. Ordinary WordNet definitions support language recognition; they are not evidence of a dream meaning. Editorial dream guides are labeled hypotheses. A source citation identifies supporting material and does not prove a personal interpretation. The optional journal AI feature described below is a separate, consent-based feature.

Optional Google Analytics

Google Analytics 4 is off by default. If you choose Allow analytics in Cookie Settings, Google may set analytics cookies and receive limited network and device information to measure selected public page visits. Our pageview events omit URL queries, fragments, search terms, private journals, interpretation pages and account pages. We do not enable advertising signals or personalization. The choice is stored in this browser and can be withdrawn in Cookie Settings; withdrawing also attempts to clear Google Analytics cookies on this site. Google processes information under its own terms and privacy policy. This browser choice is separate from the future, account-based optional usage analytics program in Preferences. Private dreams are not intentionally sent to Google Analytics.

Dream content

We do not sell dream content, publish private narratives automatically, or send raw dream text to third-party analytics. Personalized dreams and interpretations are visible only within the account unless the user explicitly shares a separate copy with a selected community audience.

Optional journal AI

When enabled, AI journal reviews and questions require your explicit session consent. Selected saved dream text, titles, dates, waking feelings, and conversation questions are sent to OpenAI for processing. Dreams on interpretation hold or excluded from patterns are omitted from retrieval. Requests disable response storage, but provider retention policies may still apply. Conversations are not saved by us and clear when you leave the screen. AI observations may be mistaken; linked quotes let you check the original account.

Separate encrypted writing

The optional encrypted-writing vault encrypts its text in your browser before storing an encrypted record. Your passphrase is not sent to us and cannot be recovered by us. Unlocked writing is present in browser memory; not all metadata is hidden. This vault does not encrypt ordinary journal entries. Copying text into a journal entry places that copy under ordinary journal storage and interpretation rules.

Illustrated diary

Diary notes, stickers, and drawings remain private and are not used as interpretation evidence. New AI illustration requests are currently paused. Existing illustrations remain private and available in your journal; there is no automatic fallback to an external image provider. Generated art may contain invented details and is labeled as AI illustration. Page exports include illustrations. Permanently deleting a dream or deleting your account removes its diary pages and stored illustrations. Moving a dream to Trash retains them until you restore or permanently delete it. Unplaced illustration previews can remain until the dream is deleted or a later page save cleans up previews older than one day.

Dream History

Saved account versions and revised interpretations remain private to your account. History begins with the account available when logging starts; it cannot reconstruct earlier edits. Moving a dream to Trash retains its history and readings until you restore or permanently delete it. Dated reading reflections are private, are included in your journal export, and do not automatically change interpretations. They are removed when you permanently delete the dream or account.

Quick-capture drafts and essential storage

Quick capture automatically keeps a draft in the current browser tab’s session storage. You may separately choose a persistent quick-capture copy in this browser; that copy has no automatic seven-day expiry. Saving the entry or disabling that option removes the persistent copy. Signing out does not currently clear that separate persistent quick-capture copy. Interpretation handoff also uses session storage. Cookie Settings lets you clear these local drafts and saved-entry recovery copies. Essential authentication cookies and browser session storage keep sign-in working. Google Analytics cookies are optional and remain off unless you allow them in Cookie Settings on this browser. ChatGPT advertising measurement is separately optional and off by default.

Recovery on your device

Recovery for saved journal entries is optional and off by default. If enabled, unsaved dream details, reflections, and illustrated pages are stored in this browser profile. These copies are not encrypted separately from browser storage; anyone with access to the profile may be able to read them. They expire after seven days and are removed when the journal next opens. Signing out or disabling recovery clears these journal recovery copies. Your last journal view and reading position are also remembered on this device and cleared on sign-out. Browser clearing or storage limits can remove recovery copies; recovery is not a backup. Downloaded copies remain under your control.

Email choices

Journal reminders and introductory marketing emails require separate opt-ins. Marketing messages do not use dream text, interpretations, or emotional profiles for targeting. When delivery is enabled, Resend receives the destination address and email content. We record consent, sending attempts, delivery feedback, and unsubscribe requests. Our templates contain no tracking pixels or tracking links. Account deletion removes account-linked marketing preferences, consent records, and queued messages; a hashed address may remain on the suppression list to prevent further unwanted mail.

Access, export and deletion

You can delete individual records, export your journal and use account-deletion controls. Trash retains a dream until you restore or permanently delete it. Account deletion removes associated private records through the service’s deletion process; copies you exported or shared with others are outside that process. Separate beta applications, email correspondence and provider records may need separate handling. We may retain limited records where needed to meet legal obligations, prevent abuse, maintain email suppression or resolve a dispute. We do not promise that every provider or backup copy disappears immediately. Contact help.me@uncannypost.com about access, correction or deletion, including rights applicable in your location.

Private people and dream notifications

Names, relationship labels, aliases, and dream-person tags are private to their owner. Connecting a person to a community account requires an accepted connection. A recipient must enable dream notifications, and each notification requires the sender to confirm one of the displayed generic messages. The recipient receives the sender’s community display name and short message, without the dream account, title, dream date, private relationship label, or other tags. You can disconnect, disable incoming notifications, remove private people, and download your people and tags in Preferences. Account deletion removes these account-linked records.

Community sharing and topic contributions

Community features are currently restricted to an owner and administrator preview and are not available to members. Existing community records remain protected. Contact help.me@uncannypost.com for access or deletion requests concerning an earlier community contribution. Private journal entries are not published to the community.

Shared video posts and replies

You may record or upload a short video and deliberately publish it with a written version to the community or a selected private Circle. Camera and microphone access begins only when you choose to record. Recording previews stay in browser memory until you choose to publish. Cloudflare Stream receives new uploads directly, encodes them and delivers video and poster images under the post’s access controls. Previously uploaded clips may remain in Supabase. Video is not end-to-end encrypted. Cloudflare receives network and device information needed to deliver media when you use the player. We do not send clips for automatic transcription, AI interpretation, model training, advertising or facial recognition. Video posting does not enroll you in research or optional analytics. Share only material and appearances you have permission to share. We retain limited daily account usage counters to enforce upload and playback allowances; these expire through maintenance after the current day and two previous days. Aggregate monthly playback-authorization counts support cost monitoring. Ready but unposted uploads are queued for deletion after 24 hours. Abandoned uploads retain an in-flight reservation for up to two hours before the next cleanup pass. Deleting a post, its Circle or your account removes access to its associated video and queues the stored file for deletion. The existing hourly maintenance process deletes queued files and retries failures. People who already downloaded or shared a copy may retain it. The player closes when the tab is hidden, the account changes or its viewing session expires. Cloudflare viewing links expire within three minutes; a link already issued may work until expiry after a membership or post changes. Previously stored Supabase clips use a five-minute local playback session.

Scientific research, business analysis, and platform improvement

Social and public contributions may support future scientific research, business analysis such as understanding feature usefulness, and improvements to search, interpretation quality, and the platform. Scientific research, model training, and the separate account usage analytics program are not active. A separate optional feature accepts only structured feedback assessments that you explicitly select and confirm for product improvement. It records the selected dictionary connection, narrative role, context category, fit assessment, and reason category; dream text, names, private notes, and written corrections are not copied into that pool. Contributions remain linked internally to your account so you can download or withdraw them. Grouped results require at least 20 contributors with the same entry, role, and context, using contributions from completed weeks. They can improve follow-up questions and inform platform and business decisions; they do not establish scientific findings. Switching off product improvement excludes your contributions from future groups. Grouped summaries already saved in earlier reading versions remain there. Posting, creating a profile, joining a Circle, or contributing a map topic does not enroll you in research or grant optional analytics, marketing, or model-training permission. Optional choices default off and can be changed or withdrawn in Preferences under Privacy & consent. A saved future preference alone does not start collection. Scientific research requires a study description and a separate selection of what to contribute; external research sharing requires partner identity, scope, and separate confirmation. Private journal entries are not automatically included in these uses or added to the public corpus. Aggregate topic displays are part of the community feature you explicitly contribute to; they are not enrollment in a research study.

Service providers and security

Our hosting and storage providers process data needed to run the site; Supabase provides account and database services. Cloudflare processes shared video uploads, encoding, protected playback and responsive blog image delivery. It receives requests needed to serve those media files; private journal text is not sent as part of media delivery. Resend processes enabled service and opted-in email delivery. OpenAI receives selected content only for the separately enabled, consented journal AI feature. Providers may process data in other countries under their applicable service arrangements. We use access controls and transport security, but no service can guarantee absolute security. Do not send account credentials in support requests.

Children and policy changes

Account and community services are intended for adults aged 18 or older. Contact us if you believe a child has submitted personal information. We will update this policy as processing changes and provide notice or request additional consent when required. A future preference does not authorize an undisclosed new use of private writing.

Optional ChatGPT advertising measurement

With your separate browser permission in Cookie Settings, the OpenAI measurement pixel measures selected public-page visits and new paid memberships following checkout. It uses network/device information and first-party attribution cookies (__oppref, up to 30 days; __obref, up to 365 days). Our event payloads contain no dream text, private notes, email addresses, or account identifiers. Subscription events include the plan, price, currency, and a checkout-based event identifier to prevent duplicates. We opt events out of future user-level personalization. Turning this choice off stops future measurement and removes the pixel cookies. Google Analytics consent does not enable this measurement. A same-tab checkout marker expires for measurement after two hours. Browser restrictions or leaving checkout without returning can prevent reporting.

Questions or an accessible copy? help.me@uncannypost.com