Automatic redaction can process eligible incoming Request Video replies using the receiving workspace's selected sensitive-text categories. Zight covers detected matches with permanent opaque masks after the video has finished the required processing.
Alpha availability: Contact us to request automatic redaction. Include your workspace and explain that you want it for Request Video replies. Zight must enable the feature for your workspace before an administrator can configure it. A subscription plan alone does not enable the Alpha.
Before you start
- You need workspace administrator access to change redaction settings.
- Use the receiving workspace that owns the Request Video link.
- Automatic processing applies to eligible, completed MP4 video replies. Custom-storage videos are not supported.
- Choose the supported text categories your organization wants covered and plan who will review the processed replies.
Enable workspace redaction
- After Zight confirms access, open your workspace's Settings > Zight AI.
- Find Redaction and select Let owners redact videos in this workspace.
- Under Cover these kinds of text, select at least one category: Email addresses, Phone numbers, Card numbers, Social security numbers, AWS access keys, or API keys and other secrets.
- Select Update Redaction Settings.
These settings also allow video owners to start a scan using AI Actions > Privacy > Redact sensitive text. You do not configure automatic post-recording redaction in the Chrome blur panel.
How Request Video processing works
- Create or use a Request Video link in the enabled workspace. Follow the Request Video setup guide for collecting replies.
- The recipient records and submits their reply.
- Zight prepares the reply in the receiving workspace and starts redaction when the video is eligible, using that workspace's selected categories.
- Open the reply and confirm that redaction has finished. Review the complete processed recording before sharing it or relying on it in another workflow.
Redaction runs in the background after submission. It does not hold integration delivery until the scan finishes. A reply link can reach an integration before redaction completes, including when redaction cannot start. Do not treat an upload notification or the arrival of a reply as confirmation that it has been redacted.
If a completed reply has not been processed, its owner can use the viewer action when available. Contact Zight Support if processing fails or the action is unavailable.
Testing your own request: A reply submitted by the request owner to their own link does not start automatic redaction through this workflow. Use a different submitter when testing automatic reply processing, or run the viewer action after your own test video finishes processing.
Which copy is redacted?
When a reply comes from another workspace, Zight creates a requester-side copy. Automatic redaction applies to that copy in the receiving workspace. The responder's separate original is unchanged. Do not assume this workflow deletes or sanitizes every copy of a recording.
On the processed item, redaction replaces the current media and hides the scanned version from the normal Versions list. The masks cannot be undone by the video owner. Zight retains the scanned original privately; redaction does not delete every retained copy. Handle originals according to your organization's policy.
Scope and limitations
- Automatic Request Video processing does not mean every ordinary upload is redacted automatically. Owners start redaction manually on other eligible videos.
- Text detection can miss sensitive content or place a mask incorrectly. A no-match result is not a privacy guarantee.
- The supported categories do not include automatic face detection or general person-name detection.
- Review audio, transcripts, captions, logs, thumbnails, form answers, and other attachments separately. Visual redaction is not a guarantee that those are sanitized.
- Confirm successful processing and apply your normal access controls before sharing or forwarding a reply.
Troubleshooting
The Redaction section is missing: confirm administrator access, the selected workspace, and Alpha enablement with Zight.
A reply was not redacted: confirm that the receiving workspace has redaction enabled and categories selected, that video processing is complete, and that the item is an eligible MP4 without custom storage. A scan can complete without detecting any selected pattern.
I changed the categories: review already collected replies separately. Do not assume changing workspace settings backfills all earlier recordings.
Comments
0 comments
Article is closed for comments.