Privacy Policy
Effective Date: 2026-10-20
1. Our Approach to Privacy
FamYarn is a service operated by DSR Media LLC. “FamYarn” is a registered fictitious name (DBA) of DSR Media LLC. In this policy, “FamYarn,” “we,” “us,” and “our” refer to DSR Media LLC.
FamYarn is designed to help families preserve and share their history with care and respect. We prioritize privacy, user control, and transparency.
Your family data is private by default and only shared when you choose.
2. Information We Collect
Account Information
- Name
- Phone number (optional — used to send SMS text notifications if you opt in, and to place a telephone interview call when you or someone you invite chooses that way of recording a story). Telephone interviews are recorded and transcribed; see Section 10.
- Authentication credentials
- Billing info
Tree Data
- Individuals (living/deceased)
- Events, relationships
- Media, notes, citations
Collaboration Data
- Invitations, edits, comments
Usage Data
- Device/browser info
- IP address
- Online identifiers (e.g., cookie IDs and a pseudonymous analytics ID)
- Logs (may include limited contact data)
3. How We Use Information
- Provide service
- Enable collaboration
- Improve product
- Ensure security
- Process billing
We use analytics tools (see Section 9) for product improvement.
We also measure our advertising. While we have an active campaign on an advertising platform, its measurement pixel may run for visitors to the application’s sign-in and signup pages, including people who did not arrive through that platform’s ad. On our marketing site and other application pages, advertising measurement remains limited to visitors referred by that platform. We also report eligible conversion events from our servers. Section 9 explains the information disclosed and the browser signals that prevent disclosure. We never use your family-tree content, name, email address or phone number for advertising.
4. Sharing
Within Trees
Shared with authorized members.
Public Sharing
- Deceased persons only
- Expiring links (one year by default)
- Users responsible for content
Service Providers
Used for infrastructure, analytics, advertising measurement, email, billing, and — where you use the features that require them — transcription, artificial-intelligence processing of photographs, documents and imported records, and photo face grouping. Section 10 names these providers and states what each one receives.
Legal Requests
Only via valid legal process.
5. Living vs Deceased
A person is considered deceased based on available data (e.g., a recorded death event, or a birth year more than 120 years ago). Users must correct data if inaccurate. Living individuals are never included in public shares.
6. Contributions
Tree data is collaborative. Users are responsible for:
- Accuracy
- Privacy of living individuals
- Rights to upload content
7. Deletion & Retention
Account Data
Deleted or de-identified.
Tree Data
Remains if shared; attribution removed.
Media
Remains attached to tree.
Timing
45 days (up to 90 if needed)
Logs
Retained up to 60 days.
8. Data Access
Users may request exports (e.g., GEDCOM). Processed within 45–90 days.
9. Cookies, Pixels & Analytics
We and our analytics providers use cookies, similar technologies, and online identifiers (such as a randomly generated, pseudonymous analytics ID) to operate the service (for example, to keep you signed in) and to understand how it is used so we can improve it.
Analytics providers. We use Google Analytics and Matomo to measure and analyze usage. These tools collect information such as pages viewed, actions taken, device and browser information, and IP address, and set or read a pseudonymous analytics identifier for your browser. Google Analytics is provided by Google; see how Google uses this data at policies.google.com/technologies/partners. You can opt out of Google Analytics with Google’s browser add-on at tools.google.com/dlpage/gaoptout.
Linking analytics to your account. To understand how features are used and how people complete key actions (creating a tree, sending or accepting an invitation, subscribing), we may associate these analytics identifiers and the related usage/visitation information with your account and activity, including events we record on our own servers. We share only pseudonymous identifiers and event information with these providers — never your name, email, phone number, or family-tree content.
Advertising measurement. We advertise FamYarn on third-party advertising platforms. While a platform has an active FamYarn campaign, we may load its measurement pixel for visitors to the application’s sign-in and signup pages, whether or not the visitor selected one of that platform’s ads. The platform may therefore receive visit and conversion information about people who arrived directly or through another source. On our marketing site and other application pages, measurement remains limited to visitors referred by that platform. The information and opt-outs described below apply to both cases.
Purposes. We use this measurement to understand whether visits and conversions relate to our advertising, including when a visitor returns without an ad link, to evaluate and allocate advertising spend, and to enable platforms to improve delivery of our own advertising. We do not use it to build profiles of you for our own purposes, and we do not acquire audience or targeting lists derived from it.
Categories of information disclosed. The disclosure is limited to:
- an advertising identifier assigned by the platform itself — typically a click identifier appended by that platform to the link you selected, together with identifiers set in your browser by that platform’s pixel;
- page-visit and conversion events and their timestamps, including visits without an advertising referral, commencement of registration and creation of an account. We may add subscription and comparable product milestones, which this section already permits;
- a pseudonymous event identifier we generate so that the platform can recognize the same event reported from your browser and from our servers as one event rather than two. It is derived from your account identifier by one-way hashing and cannot be reversed by the recipient;
- technical information inherent in any web request your browser makes to that platform, including your IP address, your browser’s user-agent, the address of the page you were on, and the address of the page that referred you. We restrict the page address as described under Advertising pixels below.
Information we do not disclose. We do not disclose to any advertising platform your name, email address, telephone number, postal address, date of birth, or any Family Data as defined below — whether in identifiable, hashed, or otherwise obfuscated form. Where a platform’s documentation invites such information, we decline to provide it.
Family Data. In this policy, “Family Data” means your family tree and its contents, including the individuals recorded in it, their names, dates, relationships, biographical and event details, photographs, documents, recordings, stories, and any annotations or research you contribute. Family Data is never disclosed to advertising platforms, advertising networks, data brokers, or any party for advertising, measurement, model-training, or resale purposes. This restriction is absolute and is not subject to any setting or election.
Categories of recipients. On the application’s sign-in and signup pages, recipients may include each platform with an active FamYarn campaign, not only a platform that referred you. On our marketing site and other application pages, recipients remain limited to the referring platform. We advertise on, and may disclose the information described above to, the following platforms:
| Platform | Provider |
|---|---|
| Meta (Facebook, Instagram) | Meta Platforms, Inc. |
| Google Ads | Google LLC |
| Microsoft Advertising (Bing) | Microsoft Corporation |
| Pinterest Ads | Pinterest, Inc. |
| Reddit Ads | Reddit, Inc. |
Advertising pixels. These platforms supply a small piece of code, commonly called a pixel (also referred to as a web beacon or tag), which reports the events described above from your browser. While a platform has an active FamYarn campaign, its pixel may be loaded on the application’s sign-in and signup pages for visitors from any source. On our marketing site and other application pages it is loaded only for visitors referred by that platform and only for a limited period after that referral. It is not loaded on our blog or help site. We configure these pixels to restrict the information they collect, including by disabling optional features that would derive contact information from form fields or inspect page content, and the page address reported from pages that reference individual records is restricted to our domain.
Legal characterization and opt-out. Under certain US state privacy laws, the disclosures described in this section constitute “sharing” of personal information for cross-context behavioral advertising and may constitute “targeted advertising”. You may opt out at any time by the means described below, at no cost and without providing any information to us.
Do Not Track / Global Privacy Control — how to opt out. Where your browser transmits a Do Not Track or Global Privacy Control signal, we give effect to it automatically. No advertising pixel is loaded, no conversion information is disclosed to any advertising platform, no data is transmitted to Google Analytics, our own analytics operate in a cookieless mode that measures usage only in aggregate, and analytics identifiers are not associated with your account. Where we hold no record of a browser’s signal, we treat the visitor as having opted out.
This signal is the opt-out for the advertising measurement described above. You do not need to contact us or fill in a form. Global Privacy Control is built into some browsers and available as an extension for others.
Opting out at the advertising platform. A platform that receives the information described above may use it, together with information it already holds about you, to measure and personalize the advertising it shows you, in accordance with its own privacy policy as linked above. Separately from the browser signal described above, you may limit that use at the platform itself. Where Pinterest is the platform, you can do so through your Pinterest personalization settings at pinterest.com/settings/privacy, and through the Digital Advertising Alliance’s AdChoices tool at optout.aboutads.info, which covers each of the platforms listed above that participates in that program. These platform-side controls are in addition to, and not a substitute for, the Do Not Track or Global Privacy Control signal described above, which prevents the disclosure from being made at all.
No sale. We do not sell personal information, and we do not sell or disclose Family Data for any advertising, measurement, model-training, or resale purpose. The only advertising-related disclosure we make is the visit and conversion measurement described in this section, which is limited to the categories of information identified above and which you may switch off by means of the browser signal described above.
10. How We Process Your Content
To provide features you ask for, some of the content you add to FamYarn is processed by automated systems, including artificial-intelligence services operated by the providers named below. These providers act as our service providers, on our instructions, for the purpose described and no other.
Processing is tied to the feature. Content is sent for the processing described here when you use the feature that requires it. If a tree enables AI photo categorization, uploaded photos are classified and may be assessed for whether people appear in them even when face matching is off. This classification does not identify the people or create stored face models. Face matching is a separate process with the choices described below. Content in a tree where you use none of these features is stored and shown back to you, and nothing else.
| Provider | What it receives | Why |
|---|---|---|
| OpenAI | Photographs you upload for AI categorization; recorded interview audio and transcripts; place and event details from a family-tree file you import; memorial-page text you ask us to summarize | Describing and tagging photographs, estimating whether people appear in them, processing interviews, and interpreting imported records |
| OpenRouter | Recipe and document page images you ask us to read, and recipe text used to suggest a title | Routing these requests to Google Vertex AI |
| Google (Vertex AI, Gemini) | The recipe and document page images and recipe text routed through OpenRouter | Reading recipe and document text and suggesting recipe titles |
| AssemblyAI | Recorded interview audio | Transcription and speaker separation |
| Deepgram | Live audio during an in-person interview | Real-time captions |
| Amazon Web Services (Rekognition) | Photographs you upload, where photo face grouping is enabled for that tree | Grouping photographs by the same face — see below |
| Twilio | Telephone numbers, SMS message content, and the audio of telephone interviews | Sending text messages and placing and recording interview calls |
| SendGrid or Amazon Web Services (SES) | Your email address and the content of emails we send you | Delivering email |
| Our gift-card provider | Your email address, your first name, and an amount — nothing else | Delivering a survey reward you have chosen to claim |
| Amazon Web Services (S3), Hetzner | Stored content; the servers the service runs on | Storing your files and operating the service |
| Stripe | Payment details, which we never see in full | Taking payment |
Model training. FamYarn does not use Family Data to build, train, fine-tune or improve AI models of its own. We do not permit any of these providers to use your content to train or improve their models, and we have taken the specific steps each provider requires to give effect to that — an organizational opt-out with Amazon Web Services, and a per-request exclusion on every Deepgram call. If a provider we use will not commit to this, we will stop using them for your content rather than accept it. Recipe and document reading requests through OpenRouter are restricted to Google Vertex AI endpoints with a zero-data-retention policy. This routing restriction does not describe how long FamYarn keeps the original content or the results you save.
Agents you choose. If you choose an external AI agent to operate your account, that agent may receive information it accesses on your behalf. Its provider is your chosen service, not one of the processing providers we select in the table above. Review that provider’s data-use and retention terms before giving it access. Our permission for agent-assisted use does not authorize model training, bulk collection or contact with other people beyond the limits in our Terms of Service.
Recorded interviews. Telephone and in-person interviews are recorded and transcribed. Everyone on the call is told the call is being recorded before the interview begins. If you invite someone else to be interviewed, you are responsible for their agreement to be recorded, and recording laws in some states require the consent of everyone taking part.
Photo face grouping
Photo classification and face matching are different processes. Classification can estimate whether people appear in a photo, as described above, without identifying them or storing face models. When face matching is on and the required matching choices allow it, new photos that Photo Insights finds people in are automatically checked against faces recorded in this tree, whether or not the photos are attached to a person. You can also ask to check a photo. We ask you to confirm every suggested name.
Every suggested identity requires a member’s confirmation, even for a very close match. A suggestion never assigns a name by itself. A member may also identify someone when setting their profile photo or by choosing a name for a face. Subject to the applicable matching choices, a face model is stored only for a face identified as a person in the tree. Face models numerically represent faces and allow images to be compared. Depending on your jurisdiction, they may be treated as biometric information.
For a face that has not been identified, measurements may be used during a permitted matching check, but no face model of that face is kept afterward. Its position in the photo and an estimated age range and sex are kept so a member can name it later. A member can also keep a name-only label: the name, the face’s position, its estimated age range and sex, and the photo’s connection to the person’s page, without a stored face model. Changing the tree’s matching setting does not automatically convert a name-only label into a recorded face.
New trees start with matching on for people recorded as deceased. A current tree admin can change who the tree covers, including extending it to living people after accepting the current terms and confirming they have permission. Any editor can turn matching off. Narrowing the scope removes recorded faces for people no longer covered; photos and names members confirmed remain.
Each member answers for their own uploads when the tree covers everyone. A photo from a member who has not accepted the current face-matching terms is not matched unless they accept or another member who has accepted explicitly takes responsibility for that photo. An account-level refusal to have uploads matched cannot be overridden this way.
When a tree covers everyone, new matching pauses if the accepting admin leaves, loses admin access, or the terms change. Existing photos, recorded faces and confirmed names remain. A current admin must accept the current terms before new matching can resume. Accepting terms or changing scope does not itself start a check of earlier photos; that is a separate choice.
People, photos and accounts have separate matching choices. Excluding a person prevents storing a face model for that person or proposing their identity; it does not prevent detecting an unnamed face in another permitted photo. Excluding a photo prevents face matching of that photo. Any editor can exclude a person or a photo in a tree. If you hold an account, you can refuse recording of your own face across trees where your account is linked to you, and refuse face matching of your uploads. A tree-wide setting cannot override a person, photo or account exclusion. People who can edit a tree may see which choice prevents matching, including when the tree owner’s location answer is the reason. These settings are not private from other authorized tree members.
Face models are kept separately for each tree. A face model created in one tree is never used to match a photograph in another tree. We use the recognition service to hold the face measurements. For a recorded face, FamYarn also keeps the service’s estimated age range, gender and expression to help organize photos. These are estimates of appearance, not verified facts. Information a member confirms, such as a name, remains their contribution.
What is destroyed, and when. Turning matching off for a tree destroys its face models. A face model is also destroyed when its photo is deleted, when the person or photo is excluded, when the relevant account is deleted, or when we receive a request to destroy it. Related estimated age, gender and expression are deleted with the recorded face; photos and confirmed names remain unless separately deleted. Narrowing scope removes the recorded faces that are no longer covered. Pausing new matching because acceptance is no longer current does not itself delete existing recorded faces.
Our published deadline for deletion from active systems is 24 hours from the event that triggers it. In practice it is usually minutes: a background process runs every 15 minutes to carry it out. Encrypted database backups are used only to restore service and are deleted within 120 days of being created. We do not retain face models for any other purpose or use them to identify anyone outside their tree.
Face models and information derived from them are never sold or disclosed for advertising, data brokerage or model training. They are processed only by FamYarn and the recognition service for the purposes described here. They are Family Data as defined in Section 9.
Location restrictions and other people’s permission. Before an action would record a face, we ask whether the acting member lives in, or is currently in, Illinois, Texas or Washington. An affirmative answer prevents new face recording from that member’s actions. If the tree owner’s answer is affirmative, new face recording is prevented for that tree. Existing recorded faces are not deleted merely because a location answer changes; a member can use the removal or opt-out controls or contact us to request deletion. Members can change their own answer in account settings, and other authorized tree members may be told when the owner’s answer prevents matching.
A tree admin’s or member’s agreement does not establish consent from everyone shown in a photo. Living people can have rights over their biometric information, including when they have no FamYarn account. The location checks do not replace permissions required by applicable law. To request destruction of a face model of yourself, or of someone whose photos you can identify, contact support@famyarn.com; the published destruction deadline applies.
11. Communications
Essential emails required. Other emails are opt-in.
12. SMS & Text Messaging
If you provide a mobile phone number and opt in, FamYarn may send you SMS text messages. Providing a phone number and opting in to texts is always optional and is never a condition of using the service.
Types of messages. Depending on the opt-ins you choose, texts may include account and billing notifications (for example, payment or plan-status alerts, referral activity, and tree-invitation or account-activity notices) and, only if you separately opt in, promotional or marketing messages about new features and offers.
Message frequency varies based on your account activity and the notifications you have enabled.
Message and data rates may apply. These charges come from your mobile carrier and are your responsibility.
How to opt out. You can cancel SMS messages at any time by replying STOP to any message. After you send STOP, we will send one confirmation message and then stop sending you texts. Reply HELP for help, or contact us at support@famyarn.com.
No mobile information will be shared with third parties or affiliates for marketing or promotional purposes. We do not sell, rent, or share your mobile phone number or your SMS opt-in and consent information with any third party for their own marketing. Phone numbers and message content are shared only with the messaging service providers (such as our SMS carrier and messaging aggregator) strictly as needed to deliver the messages you have requested, and text-messaging opt-in data and consent are excluded from all other data-sharing described in this policy.
13. Children
Users must be 13+. Genealogical data may include minors.
14. Security
We use reasonable safeguards including encryption. No system is perfectly secure.
15. Inactive Accounts
Accounts inactive for 3 years may be removed.
16. Do Not Sell or Share
We do not sell personal information.
We do not sell or share Family Data. Family Data — your family tree and its contents, as defined in Section 9 — is not disclosed to advertising platforms, advertising networks, data brokers, model-training providers, or any other party for advertising, measurement, model-training, or resale purposes. This restriction is absolute and is not subject to any setting or election.
We do disclose a limited set of visit and conversion information to advertising platforms for advertising measurement, as described in Section 9. Under certain US state privacy laws this constitutes “sharing” for cross-context behavioral advertising and may constitute “targeted advertising”. The categories of information disclosed, the purposes, and the recipients are set out in that section.
On the application’s sign-in and signup pages, this can include visitors who did not select one of our ads. While a platform has an active FamYarn campaign, its pixel may receive the limited visit and conversion information described in Section 9. Our marketing site and other application pages remain limited to visitors referred by that platform. The same opt-out applies in every case.
To opt out, enable Global Privacy Control in your browser. We honor it automatically: no advertising pixel is loaded and no conversion information is disclosed to any platform. A Do Not Track signal is treated the same way. No request to us is necessary, and we do not require an account or any information from you in order to give effect to it.
Your mobile information and SMS consent data are excluded from all advertising-related disclosure. FamYarn does not share your mobile information or SMS consent with third parties or affiliates for marketing or promotional purposes.
17. United States
Service operated in Missouri, USA.
18. Changes
We may update this policy and notify users.
19. Contact
DSR Media LLC (dba FamYarn), 701 Market St Ste 110 PMB 1942, Saint Louis, MO 63101.
Email: support@famyarn.com