Legal

Privacy Policy

Last updated: 2 July 2026

We built Earwig for journalists, documentary makers, and researchers. Your audio, your transcripts, your data belong to you. We process what we need to deliver the service and nothing more. When you record a shoot, we store the session audio (and, if you use those features, video and subtitle files) on our infrastructure so we can transcribe it, keep the audio and footage in sync, and let you play it back. You control how long we keep it, and you can delete it at any time.

Who we are

Earwig is run by Harry Winteringham (sole trader, trading as HW Studio), based in the United Kingdom. For UK GDPR purposes, Harry Winteringham is the data controller.

Email: support@earwig.io

What we collect

  • Email address — for login and account identification
  • Google account (optional sign-in) — if you choose to sign in with Google, we receive your name, email address and profile identifier from Google to create and identify your account. We do not receive your Google password.
  • Session data — transcript text during live sessions, fact-check results, and the markers, notes and labels you log while shooting. Transcript text is encrypted at rest using AES-256-GCM with a unique key per user. Only you can decrypt your transcripts.
  • Audio recordings — when you record a shoot or session, the audio is captured and stored on our infrastructure (Cloudflare R2) tied to that session. We use it to produce the transcript, to keep your audio and footage in sync, and to let you play the recording back. You can delete it at any time, and you can switch on automatic deletion (see Data retention).
  • Transcripts — the transcript we generate is stored with your session as text together with word-level timings and, where available, speaker labels.
  • Recording and sync data — to line the transcript up with your footage we store a small amount of timing information derived from the recording: an audio-anchor timestamp (your device wall-clock time at the moment recording started), any manual timecode offset you set, and the "stream position" in seconds for each marker or word. This is technical timing data, not a separate recording.
  • Video and subtitle files — if you use features that upload footage or generate subtitles, the video and subtitle (caption) files are stored on our infrastructure (Cloudflare R2) tied to your session so you can play them back, sync them and export them. If you only log markers on-device and never upload footage, no video is stored on our servers. Video and subtitle files are subject to the same retention controls as your audio (see Data retention).
  • Earwig Producer content — the idea or brief you type, and any structure, lines or captions Earwig generates for you, are stored with your account so you can keep working on your video.
  • Usage analytics — which features you use and session duration

Audio processing

Your recording is used in two ways:

  • Live transcription. During a session your audio is streamed to Speechmatics (a UK-based provider, ICO registered ZA137980) for real-time speech-to-text. Speechmatics processes the streamed audio ephemerally, does not retain it after transcription, and does not use your audio to train AI models by default.
  • Storage for transcription, sync and playback. Earwig also stores a copy of the session recording on our infrastructure (Cloudflare R2) tied to your session. We keep it so we can generate and re-align the transcript, keep the audio in sync with your footage, and let you play the recording back. This is separate from Speechmatics: Speechmatics does not keep your audio, but Earwig does, until you delete it or it is automatically deleted under your retention setting.

Stored recordings are held on Cloudflare R2 and are tied to your authenticated account. They are not part of any public or browsable index. You can delete a recording at any time by deleting the session, and you can enable automatic deletion after 30 days (see Data retention).

AI processing

  • Transcript text is sent to Anthropic (Claude AI) for fact-checking, research, and follow-up suggestions
  • When you use Earwig Producer, the idea and text you give it are sent to Anthropic to plan your video, suggest your lines and write captions
  • AI-generated results (fact-checks, research articles) are stored with your session
  • Anthropic does not use API data to train AI models. A Data Processing Addendum with Standard Contractual Clauses covers international data transfers

Special category data

If you use Earwig to record interviews or conversations in medical, clinical, or emergency services contexts, the audio and transcripts may contain special category personal data about third parties — specifically health data. Under UK GDPR, this data carries the highest level of protection.

If you are recording in a context involving patient or subject data:

  • You are responsible for ensuring that your use of Earwig complies with your organisation's data protection policies and editorial consent protocols
  • You should not push session content to shared team access until consent has been confirmed by the relevant subject
  • We strongly recommend using Earwig's consent-gated workflow which holds transcripts on your device until consent is confirmed before transmitting to our servers

We do not knowingly store special category data beyond what is necessary to provide the service. If you become aware that a session contains health data relating to an identifiable individual who has not consented to recording, please delete the session immediately and contact us at support@earwig.io.

Live Share and producer sharing

When you share a session using a Live Share PIN, real-time transcript content becomes visible to the people you share it with. You are responsible for ensuring that the people you share sessions with are authorised to receive that content.

Live Share sessions are not stored separately — they mirror the content of your active session. When your session ends, the Live Share connection closes. Recipients of a Live Share session do not retain a copy on our servers unless they have their own Earwig account and explicitly save the session.

Competitions and content you submit

If you enter one of our competitions by sharing content you have made or logged with Earwig, we collect the video or files you send us and the contact details you provide (such as your name and email or social handle).

We use this only to run the competition — reviewing entries, and choosing and contacting winners — and, where you have given us permission, to feature your entry with credit in our marketing. The legal basis is your consent, which you can withdraw at any time by emailing support@earwig.io.

We keep entries for as long as we need them to run the competition, and, if you have agreed to marketing use, for as long as we feature them or until you ask us to stop. You can ask us to delete your entry at any time.

The /questions feature

When you use Earwig's question preparation feature (/questions), the following data is processed:

  • Question sets — stored in memory during your session (not persisted to our servers in v1)
  • Session goal — if provided, sent to Anthropic alongside your questions for AI analysis
  • Question status — answered, flagged, or skipped status tracked during your session
  • Pickup reports — generated by Anthropic based on your question outcomes and goal

Your question sets and session goals are sent to Anthropic (our AI provider) to generate question suggestions and pickup report analysis. Anthropic does not use API data to train AI models.

Third-party services

Third party

Speechmatics — real-time speech-to-text

Audio is streamed to Speechmatics for live transcription. Speechmatics is a UK-based company (Cambridge), ICO registered (ZA137980), certified to ISO 27001:2022 and SOC 2 Type II. Audio is processed ephemerally and discarded immediately after transcription. Speechmatics does not use customer audio to train AI models by default.

speechmatics.com/legal/privacy-policy
Third party

Anthropic (Claude AI) -- fact-checking and research

Transcript text is sent to Anthropic's Claude API for real-time fact-checking, research, and follow-up question generation. Anthropic does not use API data to train AI models. A Data Processing Addendum with Standard Contractual Clauses covers international data transfers.

privacy.claude.com
Infrastructure

Render — application hosting

Application hosting on EU/US servers. Your data is processed within Render's infrastructure. Transcript data is encrypted at rest before being written to the database — Render infrastructure does not have access to plaintext transcript content.

render.com/privacy
Infrastructure

Cloudflare R2 — media storage

Your session audio recordings, and any uploaded video and generated subtitle files, are stored as objects on Cloudflare R2. Files are tied to your account and are accessed through short-lived, authenticated links; there is no public or browsable index of your media. Files are deleted when you delete the session or account, or automatically if you enable footage auto-delete.

cloudflare.com/privacypolicy
Third party

Resend — login emails only

Used exclusively for sending authentication and login emails. No marketing communications.

resend.com/legal/privacy-policy
Third party

Google — optional sign-in

If you choose "Sign in with Google", Google authenticates you and shares your name, email address and profile identifier with us so we can create or access your account. We do not receive your Google password and we request only basic profile and email scopes. Sign-in is optional; you can instead use an email login.

policies.google.com/privacy
Third party

Stripe — web payment processing

On the web, subscription payments are processed by Stripe. We do not store, access, or process your payment card details — all card data is handled directly by Stripe's PCI-DSS compliant infrastructure. Stripe may store your email address and payment method for billing purposes.

stripe.com/privacy
Third party

Apple — App Store subscriptions

On the iOS app, subscriptions are purchased through Apple's in-app purchase. Apple processes the payment and manages billing; we receive only your subscription status, never your payment details. Governed by Apple's privacy policy.

apple.com/legal/privacy
Third party

PostHog — product analytics

We use PostHog (EU-hosted) for privacy-friendly product analytics — which features are used and aggregate usage trends — to improve Earwig. Not used for advertising or cross-site tracking, and we do not sell this data.

posthog.com/privacy
Third party

Instagram / Meta — optional sharing

If you choose to share a video you made with Earwig Producer to Instagram, that content and any related information is handled by Instagram (Meta) under their own policies. We only share to Instagram when you ask us to. Earwig is not affiliated with or endorsed by Meta.

privacy.instagram.com

Data retention

  • Streamed audio at Speechmatics — processed ephemerally and discarded immediately after transcription; not retained by Speechmatics
  • Stored audio recordings (Cloudflare R2) — retained until you delete the session or your account. If you enable footage auto-delete, recordings and associated media are automatically deleted after your chosen retention period (30 days by default). Deleting your account removes them.
  • Video and subtitle files (Cloudflare R2) — retained on the same basis as your audio recordings, including your footage auto-delete setting
  • Transcripts, word timings and speaker labels — retained until you delete the session or your account. Deleting your account (in-app or on request) removes them immediately
  • Recording and sync data (audio-anchor timestamp, timecode offset, marker stream positions) — retained with the session it belongs to and deleted with it
  • Question sets — retained in memory during session only (v1). Not persisted to database
  • Pickup reports — retained in memory during session only (v1)
  • Live Share session data — not stored separately. Exists only during the active session
  • Account email — retained until you close your account or ask us to remove it
  • Consent records — retained for audit purposes as required by law
  • Server logs — 30-day rolling deletion

We never sell your data.

Your rights

  • Access your data
  • Delete your account and all associated data — instantly, in the app (Marker → profile → Delete account), or by emailing us. This removes your stored audio recordings, video and subtitle files, transcripts and sessions.
  • Delete individual sessions, which removes their audio, video, subtitle files and transcript
  • Turn on footage auto-delete so recordings and media are removed automatically after your chosen retention period
  • Export your transcripts and fact-check results

To exercise any of these rights, contact support@earwig.io

Cookies & local storage

Earwig only uses storage that is strictly necessary to provide the service you signed up for. Under PECR (the UK ePrivacy regulations) these categories are exempt from cookie-consent requirements, so we don’t show a consent banner.

  • Session cookie (sid) — required for authentication. HttpOnly, SameSite=Lax, served over HTTPS in production.
  • Browser local storage — saves your preferences inside the app (recording mode, Whisper voice, transcription language, AI feature toggles, prepped questions, recent UI state). All keys are prefixed earwig_. Nothing is shared with third parties; clearing your browser storage resets these to defaults.
  • Stripe cookies during checkout — if you upgrade your plan, Stripe sets its own cookies on its hosted checkout page to process the transaction. They’re essential to complete the payment and live on Stripe’s domain, not Earwig’s.
  • Product analytics (PostHog) — we use PostHog (EU-hosted) to understand which features are used and improve the app. It is not used for advertising or cross-site tracking, and we do not sell this data.
  • No third-party advertising cookies. No cross-site tracking.

We do not show third-party ads. If we ever add anything that genuinely requires consent, we’ll add a proper banner before turning it on.

Legal basis for processing (GDPR)

Under UK and EU GDPR, we process your personal data on the following legal bases:

  • Contract (Art. 6(1)(b)) — processing your email, session data, audio recordings, video and subtitle files, transcripts and the associated timing/sync data is necessary to provide the recording, transcription, sync and playback features you signed up for
  • Legitimate interest (Art. 6(1)(f)) — usage analytics to improve the service and ensure security, balanced against your privacy rights
  • Consent (Art. 6(1)(a)) — AI-powered features (fact-checking, research, questions) are optional and can be disabled at any time from your settings

You have the right to: access your data, rectify inaccuracies, erase your data, restrict processing, data portability, and object to processing. To exercise any right, email support@earwig.io. We will respond within 30 days.

If you believe your data rights have been violated, you may lodge a complaint with the Information Commissioner's Office (ICO) at ico.org.uk.

California privacy rights (CCPA)

If you are a California resident, you have the following rights under the California Consumer Privacy Act:

  • Right to know — you can request the categories and specific pieces of personal information we have collected about you
  • Right to delete — you can request deletion of your personal information
  • Right to opt out of sale — we do not sell your personal information to third parties
  • Right to non-discrimination — we will not discriminate against you for exercising your privacy rights

Categories of data collected: identifiers (email address, Google profile identifier where you sign in with Google); audio and visual information (session audio recordings, and uploaded video and generated subtitle files); internet or other electronic activity (session data, markers, usage analytics, recording timing/sync data); professional information (transcript content). We collect this data to provide the Earwig service. We do not sell or share (as those terms are defined under California law) your personal information. Third-party processors (Speechmatics, Anthropic, Cloudflare, Google, Render) receive data only as needed to deliver service functionality.

To make a CCPA request, email support@earwig.io. We will verify your identity and respond within 45 days.

International data transfers

Your data may be transferred to and processed in countries outside the UK/EEA. Where this occurs, we ensure appropriate safeguards are in place:

  • Speechmatics — UK-based (Cambridge), processes data within UK/EU infrastructure
  • Anthropic — US-based, covered by a Data Processing Addendum with Standard Contractual Clauses (SCCs)
  • Cloudflare (R2) — stores your audio, video and subtitle files on Cloudflare's global network; data may be processed in the United States and other regions. International transfers are covered by Standard Contractual Clauses.
  • Google — US-based, used only for optional sign-in; transfers covered by Standard Contractual Clauses
  • Render — hosting infrastructure in EU/US regions
  • Resend — US-based, used only for authentication emails

How to complain

If you have any concerns about our use of your personal data, please contact us first at support@earwig.io.

If you remain unhappy after raising a complaint with us, you can complain to the ICO:

Information Commissioner's Office
Wycliffe House, Water Lane
Wilmslow, Cheshire SK9 5AF
Helpline: 0303 123 1113
ico.org.uk/make-a-complaint

Frequently asked questions

Can Earwig access my transcripts?
No. All transcripts are encrypted at rest using AES-256-GCM with a unique encryption key generated for each user. This means the transcript text is scrambled before it is ever written to the database. Earwig staff, our hosting provider (Render), and any third party with access to the database would see only unreadable ciphertext — never the words spoken in your interview. Only your own account can decrypt and read your transcripts.
Is my audio stored?
Yes. When you record a shoot or session, Earwig stores the audio recording on our infrastructure (Cloudflare R2), tied to your account, so we can generate and re-align the transcript, keep the audio in sync with your footage, and let you play it back. This is separate from the live transcription: while you are recording, the audio is also streamed to Speechmatics, which transcribes it and discards it immediately without keeping a copy. Your stored recording stays until you delete the session or account, or, if you turn on footage auto-delete, until it is automatically removed (30 days by default). If you only log markers on-device and never record or upload, no audio is stored on our servers.
Does Anthropic see my transcript content?
Only when AI features are active (fact-checks, questions, research, Pickup Reports). When you enable AI, excerpts of the transcript are sent to Anthropic's Claude API to generate suggestions. Anthropic does not use API data to train AI models, and this is governed by a Data Processing Addendum. You can turn AI off at any time from your account settings — when it is off, nothing is sent to Anthropic.
Who can see my sessions in the gallery?
Only you. Sessions are tied to your authenticated account and are not accessible to other users. Share links, when generated, use a token-based URL — anyone with the link can view a read-only version of that session, but no browsable index of your sessions is ever exposed publicly.
Can I delete my data?
Yes. You can delete individual sessions directly from the gallery. To request full account deletion and removal of all associated data, email support@earwig.io and we will process it within 30 days in line with your rights under UK GDPR.

Your responsibilities as an Earwig user

When you use Earwig to record and transcribe a conversation, the person you are interviewing is a data subject under UK GDPR. You are responsible for ensuring they have given their informed consent before recording begins.

By starting a session on Earwig, you confirm that:

Earwig provides a built-in consent screen to help you meet this obligation. We strongly recommend using it for every session.

If you are recording in a medical, clinical, or emergency services context, additional obligations may apply. Please consult your organisation's data protection policy before using Earwig in these environments.

Questions about your data?

If you have any questions about how we handle your data or want to make a request, get in touch.

support@earwig.io