Human Archive

Privacy Policy

Effective date: July 30, 2026

1. Introduction

Human Archive, Inc., a Delaware corporation (“Human Archive,” “we,” “us,” or “our”), operates the Human Archive mobile application and related contributor services (collectively, the “Service”). The Service lets adult contributors record first-person video of everyday tasks and household activities, submit those recordings to us, and earn payouts for submissions that we approve. We then process those recordings and license the resulting datasets to artificial-intelligence and robotics customers to help improve embodied AI systems.

This Privacy Policy describes the personal information we collect from and about contributors through the Service, how we use and share that information, and the choices and rights available to you. By creating a contributor account or otherwise using the Service, you acknowledge that you have read and understood this Privacy Policy.

2. Service Is for Adults Only

The Service is intended only for individuals who are at least 18 years old. The Service is not directed to, and we do not knowingly collect personal information from, anyone under 18. You must confirm that you are at least 18 when you create a contributor account, and you must not allow anyone under 18 to use your account or appear as the primary recording subject in any submission.

If you become aware that a person under 18 has provided personal information to us, please contact us at the address in Section 13 below and we will take steps to delete that information and terminate any associated account.

3. Recording in Homes and Businesses; Third Parties Captured in Recordings

The Service offers two kinds of contributor tasks: video tasks (first-person video of everyday tasks in homes and businesses, with no audio captured) and Voice Exchange tasks (voice recordings captured through the microphone, with no video captured). Because video tasks record in homes and businesses, your video recordings may capture other people, their property, identifying documents, screens, signage, addresses, license plates, and similar information. Because Voice Exchange conversations record audio from you and, in two-party conversations, another Human Archive contributor, voice recordings may capture what you or the other contributor say aloud, including any information about other people or sensitive information that you choose to discuss. You are responsible for following our task-specific rules and any instructions we provide about excluding bystanders, minors, sensitive surroundings, and prohibited content, and, for Voice Exchange conversations, for not sharing sensitive personal information (see our Contributor Recording Rules). We may reject, redact, or delete any submission that contains content we determine is non-compliant.

Before each recording session, we present a consent confirmation that you must affirmatively accept. We log that confirmation, along with the task you selected and basic session metadata, as part of the Submission and QA information described below.

4. Information We Collect

We collect the following categories of personal information from and about contributors. Where we describe categories below, the examples we give are illustrative, not exhaustive.

4.1 Information you provide to us

Personal information you may provide to us through the Service or otherwise includes:

Account and profile information — for example, your name, email address, contributor ID, country or region, profile preferences, and account status.
Communications data — for example, the contents of your messages to us through in-app support, email, surveys, or any other communications channel we make available.
Recording and session information (video tasks) — for example, the first-person video you record, depth files, IMU and motion data, camera-calibration data, and associated session metadata. Metadata includes information about how, when, and where a recording was collected, the device used to collect it, the task or clip identifier, duration, timestamps, and any keywords, region, or other contextual information added to or associated with the recording. Video tasks capture video and motion data only — no audio is recorded during a video task. Audio is collected only during Voice Exchange tasks, described in Section 4.4.
Submission and QA information — for example, the task you selected, the consent confirmations you provide before recording, clip review status, quality-check results, approval and rejection decisions, any reversal or revocation of an approval after initial review, your submission history, and any compliance flags.
Payment and payout information — for example, your earnings balance, payout history, selected payout method, payout status, country, and identifiers issued by our payout provider. Depending on the payout provider you use, you may also provide additional payment information directly to that provider under its own terms and privacy policy.
Government-issued identification information — for example, your Social Security or other tax-identification number, driver’s license or state identification number, passport number, and (where required) an image of the identification document. Where required for payouts, you provide this information directly to our payout provider under its own terms and privacy policy; we receive only the status of the identity-verification process and not the underlying identification information itself.
Other data — other information that you provide to us or that we collect as described in this Privacy Policy or as we describe to you at the time of collection.

4.2 Information from third-party sources

We may combine the personal information we receive from you with information falling within one of the categories above that we receive from third parties, such as:

Payout providers — who may share back account-status, identity-verification status, tax-status, or transaction-status information so that we can manage eligibility and reconcile payouts.
Fraud, security, and identity-verification vendors — who may share signals indicating that a sign-up, login, or submission warrants additional review.
Platform providers — for example, Apple and Google may share information that they collect as the platform operators (such as crash diagnostics) and that we use to operate the app.
Business-transaction counterparties — we may receive personal information in connection with an actual or prospective merger, acquisition, financing, or similar transaction described in Section 6.

4.3 Information we collect automatically

We, our service providers, and our platform partners may automatically log information about you, your device, and your interaction over time with the Service, such as:

Device and capability data — for example, your device model, operating system version, app version, and the availability of your camera, LIDAR/depth sensor, motion sensors, and other relevant device capabilities.
Location and region data — for example, the country or region you provide and other location information you choose to share when you grant the location permission on your device.
App-activity data — for example, the screens you view, the actions you take, the time you spend on a screen, and the navigation paths you follow within the app.
Support, security, and diagnostics data — for example, crash logs, device and app diagnostics, audit logs, IP address and network metadata, and security events.

The third-party software-development kits (SDKs) currently embedded in the app are limited to (i) the Supabase SDK, which supports our authentication, contributor database, and backend services, and (ii) MediaPipe, which performs certain on-device image analysis used to calculate aggregate quality metrics about your recording (for example, the percentage of frames containing a hand). MediaPipe runs entirely on your device, and neither the recordings, the frames, nor the underlying biometric landmarks it processes are transmitted off your device by MediaPipe.

In addition, as part of the quality-assurance review we perform on your submissions, we send selected frames from your recording to a third-party generative-AI service (currently Google’s Gemini API) to evaluate eligibility, detect prohibited content, and identify potential privacy issues. The terms under which we currently access that service permit the service provider to use submitted data, including the frames we transmit, to improve its own products and services. We do not send your recordings to Gemini for any other purpose, and we do not send your account profile, payment, or government-issued identification information to Gemini. For the Voice Exchange proficiency assessment described in Section 4.4, we also send the assessment audio clip to Google’s Gemini API for language-proficiency scoring, and this transmission is subject to the same terms as the video frame sample described in this paragraph. We do not send live Voice Exchange conversation audio to Gemini; conversation audio is routed and stored as described in Section 4.4.

4.4 Voice Exchange information (voice tasks)

Voice Exchange is a separate, optional feature of the Service in which you complete a one-time reading assessment (used to check your language proficiency before your first Voice Exchange task in a language) and take part in solo voice tasks and live two-party voice conversations with another Human Archive contributor. Voice Exchange tasks capture audio only. Video is not captured during any Voice Exchange task. Before you take part in any Voice Exchange task, we present a voice-consent screen that describes what we collect and how we use it, and before every two-party conversation we present a per-call consent screen that both participants must affirmatively accept before the recording begins.

For Voice Exchange tasks, we collect and process:

Voice recordings — including the audio you record for a solo Voice Exchange task, the audio of your reading assessment, and the audio of your live two-party conversations. In a two-party conversation, we record both participants’ voices.
Derived voice data — information derived from your voice recordings by our automated review, including a transcript of what was said, the detected language of the recording, a speaker-count value (how many distinct speakers were audible, but not who), audio-quality scores, and a probability score indicating whether the recording is likely to be AI-generated rather than a natural human voice. We do not extract voiceprints or speaker embeddings from your recordings, and we do not use your voice to uniquely identify you.
Voice-task session data — the task or conversation identifier, timestamps, duration, and the consents you provided before recording, along with technical logs from the real-time-communications provider that carried the call.

We use voice recordings and derived voice data to operate the Voice Exchange feature, to run automated quality-assurance analysis (including transcription, language check, speaker-count check, audio-quality scoring, and AI-generated-voice detection), to verify language proficiency, to develop and improve speech and embodied-AI datasets, and to license approved voice datasets to AI and robotics customers as described in Section 6. Voice datasets are licensed separately from video datasets.

Voice Exchange audio is handled by three categories of service provider on our behalf: (i) a real-time communications provider (currently LiveKit) that carries the live conversation audio between the two participants and records the conversation on our behalf; (ii) our cloud infrastructure provider (currently Amazon Web Services), which stores the resulting recordings and runs the automated quality-assurance analysis; and (iii) a third-party generative-AI service (currently Google’s Gemini API), which scores the reading-assessment audio clip only. Live two-party conversation audio is not sent to the Gemini API.

Two-party conversations are recorded on an all-party-consent basis: both participants complete the per-call consent screen in the app before the conversation begins, and no recording occurs unless both participants have consented. You can withdraw your consent to future Voice Exchange participation at any time in Settings; withdrawing consent does not remove recordings you have already submitted, except as described in Section 12.

Voice Exchange is also governed by the Voice Contributor Rules and by Section 5 of the Contributor Recording Rules.

5. How We Use Personal Information

We use the personal information described above for the following purposes:

5.1 Providing the Service

Creating and managing your contributor account and profile.
Confirming that you meet eligibility requirements, including age, region, and device capability.
Matching you with available recording tasks and, for Voice Exchange two-party conversations, matching you with another Human Archive contributor for a conversation (or connecting you with a contributor you invited by phone number).
Receiving, storing, processing, and reviewing your recording and session submissions.
Calculating pending and approved earnings and processing payouts to you.
Licensing approved recording datasets and associated metadata to AI and robotics customers, as described in Section 6.

5.2 Personalizing your experience

We may use your personal information to understand your needs and preferences, to remember selections such as preferred language, task filters, and notification settings, and to personalize your experience with the Service and our communications with you.

5.3 Operating, securing, and improving the Service

Operating, debugging, and maintaining the reliability of the Service.
Monitoring and protecting the security and integrity of our systems.
Analyzing how contributors use the Service to improve features, workflows, and dataset quality.
Performing internal research and analytics related to the Service.

5.4 Quality assurance and content compliance

Running automated and manual quality-assurance review of submissions, including, for Voice Exchange, transcription, language-detection, speaker-count, audio-quality, and AI-generated-voice checks.
Detecting and preventing prohibited content, privacy issues, fraud, duplicate submissions, and low-quality data.
Re-examining a submission after it has been approved, and reversing or revoking that approval, where a technical error, fraud, duplicate submission, or material quality or policy issue comes to light after initial review.
Enforcing our content rules, recording rules, and other policies.

5.5 Compliance, safety, and protection

Satisfying tax, KYC, sanctions-screening, anti-fraud, and other legal or regulatory obligations applicable to us or to our payout providers.
Investigating and addressing suspected fraud, abuse, security incidents, or violations of our terms.
Establishing, exercising, or defending legal claims and responding to legal process.
Protecting the rights, property, and safety of Human Archive, our contributors, our customers, and others.

5.6 Communications

Providing customer support and responding to your questions and requests.
Sending you transactional messages about the Service, such as task availability, submission status, payout status, security alerts, and updates to this Privacy Policy.

5.7 Aggregated, de-identified, and anonymized information

We may create aggregated, de-identified, or anonymized information from the personal information we collect, by removing the information that makes it identifiable to you. We may use this aggregated, de-identified, or anonymized information for any lawful purpose, including analyzing and improving the Service, performing internal research, and sharing with third parties for these purposes. We do not attempt to re-identify aggregated, de-identified, or anonymized information, and we contractually require recipients of de-identified information not to attempt re-identification.

5.8 Further uses

In some cases, we may use your personal information for further purposes that are compatible with the purposes for which we originally collected it, or for further purposes for which we will obtain your consent if your consent is required by law.

6. How We Share Personal Information

6.1 AI and robotics customers (licensing of approved submissions)

A core purpose of the Service is to license approved recording datasets to AI and robotics customers. When we approve a submission, we may include the recording itself (first-person video, any audio, depth files, IMU and motion data, camera-calibration data, and associated metadata) in datasets that we license to AI and robotics labs and similar customers under a written agreement. Those agreements restrict the customer’s use of the data to the purposes specified in the agreement, prohibit re-identification of contributors except as expressly permitted, and require the customer to keep the data confidential and secure. We do not license your account profile, payment, or government-issued identification information to AI or robotics customers. Voice Exchange recordings and derived voice data are licensed as separate datasets from video recordings; when we license a voice dataset, we include the voice recording and derived voice data (transcripts, detected language, speaker count, audio quality, and AI-generated-voice probability), but not your account profile, payment, or government-issued identification information.

6.2 Service providers and processors

We share personal information with vendors that provide services on our behalf, under written contracts that restrict their use of personal information to providing services to us. We currently use, or expect to use, vendors in the following categories:

Authentication, contributor database, and backend services — used to manage accounts, sign-in, and our core backend.
Cloud storage and supporting infrastructure — used to store uploaded recordings and related files and to host our systems.
Real-time communications services — used to carry and record live audio between the two participants in a Voice Exchange two-party conversation. Our current real-time communications provider is LiveKit.
Mobile-platform services — used for app distribution, device permissions, crash diagnostics, and other platform-level features provided by the operating-system vendors.
Payout, tax/KYC, and sanctions-screening services — used to process payouts to contributors and to satisfy applicable identity, tax, and anti-fraud obligations. Our payout providers are Stripe and Circle; certain payment, identity, or tax information you provide is collected by those providers directly under their own terms and privacy policies (Stripe’s is available at https://stripe.com/privacy).
Submission review and quality-assurance tools — used to evaluate submissions through automated review and, where applicable, through review platforms used by our trained reviewers.
Analytics, security, fraud-detection, and reliability vendors — used to operate, secure, and improve the Service.

We require each service provider to use personal information only to provide services to us and in accordance with our instructions, to maintain appropriate confidentiality and security measures, and to comply with applicable law.

6.3 Compliance, safety, and legal process

We may share personal information when we believe disclosure is necessary or appropriate to: comply with applicable law, legal process, or lawful requests from public authorities; enforce our terms; detect, prevent, or address fraud, security or technical issues; or protect the rights, property, or safety of Human Archive, our contributors, our customers, or others.

6.4 Business transfers

If Human Archive is involved in a merger, acquisition, financing, reorganization, bankruptcy, sale of all or a portion of its assets, or similar transaction, personal information may be transferred as part of that transaction. We will require the recipient to honor the commitments in this Privacy Policy with respect to personal information transferred, or we will provide notice and an opportunity to object before any materially different use occurs.

6.5 With your direction

We may share personal information with other third parties when you direct or authorize us to do so — for example, when you choose to connect your Human Archive account to a third-party service.

6.6 De-identified and aggregated information

As described in Section 5.7, we may share aggregated, de-identified, or anonymized information that does not identify you.

7. Device Permissions

The Service uses certain device permissions to function. We request permission only when we need it, and you can change your permission choices at any time in your device settings:

Camera — required to record the first-person video that is the core of the Service.
Microphone — used only during Voice Exchange tasks (the reading assessment, solo voice tasks, and two-party voice conversations described in Section 4.4) to capture and record your voice. The microphone is not used during video tasks, and video tasks do not capture audio. You can decline microphone permission and complete video tasks without granting it; you will not be able to complete Voice Exchange tasks unless you grant microphone permission.
LIDAR/depth and motion sensors — used, where your device supports them, to capture depth and motion data for tasks that require it.
Location — used to determine regional eligibility and to support task matching when you grant the location permission.
Notifications — used to notify you about task availability, submission status, and account or payout updates.

If you revoke a required permission, certain features of the Service may not function and you may not be able to complete certain tasks.

8. Other Sites and Services

The Service may contain links to websites, mobile applications, and other online services operated by third parties (for example, to the privacy policy or support resources of our payout provider, or to platform pages operated by Apple or Google). These links are not an endorsement of, or a representation that we are affiliated with, any third party, and we do not control or take responsibility for the practices of third-party services. We encourage you to read the privacy policies and other terms of any third-party website, application, or service you use.

9. Data Retention

We retain personal information for as long as needed to provide the Service to you, to operate our business (including paying contributors, calculating earnings, performing fraud and quality reviews, and meeting tax and accounting obligations), and to comply with our legal obligations. In general, we retain account and profile information while your account is active and for a reasonable period after account closure to handle final payouts, tax reporting, dispute resolution, and recordkeeping required by law. We retain rejected submissions only for the limited time needed to administer the review and any appeal, and we retain approved submissions and the metadata required to support our licensing of those submissions for the duration of the applicable customer license and our recordkeeping obligations. We retain payment, tax, and KYC information for the periods required by tax, anti-money-laundering, and other applicable laws, and we retain support, diagnostic, and security logs for the periods needed to operate and secure the Service. When personal information is no longer needed for the purposes described in this Privacy Policy, we will delete, de-identify, or archive it in accordance with our internal retention practices.

10. Security

We use administrative, technical, and physical safeguards designed to protect personal information from unauthorized access, use, alteration, and disclosure. These include access controls, encryption in transit, encryption at rest for stored recordings, logging and monitoring, and contractual protections with our service providers. No security program is perfect, however, and we cannot guarantee the security of personal information.

11. International Locations

Human Archive is based in the United States, and we operate the Service primarily from the United States. We and the vendors we use to provide the Service may store and process personal information in the United States and in other countries where we or those vendors operate. Privacy and data-protection laws in those countries may be different from, and may be less protective than, the laws of your state, province, or country. By using the Service or providing personal information to us, you understand that your personal information may be processed in those locations.

12. Your Choices and Account Controls

12.1 Account settings and consent withdrawal

You can review and update your profile information, change your communication preferences, manage device permissions, and control whether you submit new recordings at any time from within the app or your device settings. You can withdraw consent to participate in the Service at any time by stopping use of the Service and, if you choose, deleting your account.

12.2 Account and data deletion

You can delete your contributor account at any time from within the app, under Settings → Account → Delete Account, or by contacting us at the address in Section 13. When you delete your account, we will delete or de-identify personal information associated with the account, except for information that we are required or permitted to retain, including:

Recordings and metadata that we have already approved and incorporated into datasets licensed to AI or robotics customers, which we will continue to handle in accordance with the applicable customer license and our recordkeeping obligations.
Payment, tax, KYC, and anti-fraud records that we are required to retain by law.
Records reasonably necessary to detect security incidents, investigate fraud or abuse, defend or assert legal claims, or comply with law.

We will describe in your account-deletion confirmation what we have deleted and what we are retaining and why. Deleted recordings that were never approved will be removed from our active systems on the schedule set out in our internal retention practices.

12.3 Tracking technologies and Do Not Track

We use limited analytics within the app to operate and improve the Service. We do not use the Service to display third-party advertising to contributors or to share personal information with advertising networks for cross-context behavioral advertising. We do not currently respond to “Do Not Track” signals from web browsers because there is no industry-standard interpretation of those signals.

13. How to Contact Us and Submit Requests

You can ask us questions about this Privacy Policy, ask us to access or correct your personal information, or ask us to delete your personal information by contacting us at:

Human Archive — Privacy
466 Brannan Street, San Francisco, CA, USA 94107
Email: team@humanarchive.ai

We will respond to verifiable requests within a reasonable time. We may need to verify your identity before responding, and we may decline a request to the extent permitted by applicable law (for example, where granting the request would compromise the privacy of others, our ability to detect fraud, or our compliance with a legal obligation).

14. Changes to This Privacy Policy

We may update this Privacy Policy from time to time. When we make material changes, we will notify you within the app or by another reasonable means and will update the “Effective date” at the top of this Privacy Policy. Your continued use of the Service after the updated Privacy Policy takes effect will be subject to the updated Privacy Policy. We will not apply a materially different use to personal information collected under a previous version of this Privacy Policy without giving you notice and, where required, obtaining your consent.