Privacy Mode

What does Privacy Mode disable in AetherCut?

Privacy Mode turns off analytics page views, auto captions, AI Director suggestions, AI B roll generation, AI color match, and any other feature that calls the backend. Local editing and export still work normally.

• Analytics page-views (no POST /api/analytics/page-view while toggled on).
• Auto-captions / Whisper transcription.
• AI Director suggestions.
• AI B-roll generation (Kling, Nano Banana).
• AI color match.
• Any other feature that calls our backend.

The editor itself — cuts, trims, transitions, color grading, text overlays, manual subtitles, export — continues to work normally. They never needed the network anyway.

How do I turn on Privacy Mode in AetherCut?

Click the shield icon in the top right header next to your avatar. The setting is saved in localStorage under aethercut_privacy_mode, so it stays on after reloads.

Top-right of the header, next to your avatar. Click the shield icon. State persists in localStorage under the key aethercut_privacy_mode. Survives reloads.

What it does NOT disable

• The initial page load (you need the editor bundle to use the editor).
• Google Fonts (loaded once at first paint, cached locally afterward).
• Sign-in flow if you choose to sign in (Google OAuth requires a network round-trip).
• Stripe.js on the pricing page, if you choose to upgrade.

Privacy Mode is for your editing session, not your account management.

How can I verify Privacy Mode is blocking API calls?

Open DevTools Network, turn Privacy Mode on, then reload and edit or export. You should only see static asset loads at first paint and internal media blob URLs, with zero calls to /api/ endpoints.*

Open DevTools → Network. Toggle Privacy Mode ON. Reload. Edit a clip, apply a transition, export. The only requests you should see are the static asset loads at first paint and any media blob URLs the editor creates internally. Zero calls to our /api/* endpoints.

When to use Privacy Mode and what people often misread

Privacy Mode is most useful when you want a clearly offline-style editing session after the app has loaded: rough cuts on sensitive interview footage, internal review edits, client work you do not want sent to any optional cloud feature, or any session where you know in advance that AI tools are not part of the workflow. In practice, that means you should turn it on before you start experimenting with captions, AI suggestions, or generated B-roll, because those are the features most likely to fail in ways that feel confusing if you forgot the shield is enabled. A common pattern is simple: load the editor, enable Privacy Mode, import your media, then do all trimming, arranging, text, color, subtitles, and export locally without expecting any server-backed enhancement to appear.

The most common misunderstanding is treating Privacy Mode like a full website-wide network blocker. It is narrower than that. It protects the editing session from optional outbound calls after the page has loaded, but it is not meant to hide the fact that you visited the site, signed in, or opened billing. That distinction matters when you are testing it. If you turn it on and still notice the initial app bundle, font requests, or a login redirect using the network, that does not mean the feature failed. The cleaner test is to trigger actions that would normally contact the backend and confirm that nothing new is sent when the toggle is active.

Another easy mistake is forgetting why a feature suddenly looks unavailable. If auto-captions do nothing, AI Director never returns ideas, or AI color match appears unresponsive, check the shield before assuming the editor is broken. Privacy Mode is doing exactly what it promises: it prefers no optional outbound traffic over convenience. If you need one of those tools, switch it off, use the feature, then turn it back on for the rest of your edit.