A CCTV access request needs a review of the footage relevant to the requester and the information it reveals about other people. Medianonymizer can help prepare an edited copy by processing selected identifiers in an uploaded video. The person handling the request still decides what should be disclosed.
Start by finding the relevant camera and time interval. Preserve the source according to your organisation's process, then work on a copy. A short, relevant extract is easier to review than an entire shift of footage containing unrelated activity.
Identify the requester in the footage
Establish which person and appearances are relevant to the request. A busy shop can contain similar clothing, reflections and people returning through a different camera angle. Record the cameras and intervals you intend to include so the reviewer can compare the result with the request.
The EDPB's video-device guidance addresses access rights and the rights of other people in the recording. It does not mean every request has the same disclosure outcome. Assess the request and applicable obligations rather than treating an edited video as automatic approval.
Review detected faces
Upload a supported video and inspect the detected identities. Select the people whose faces should be redacted. If the requester is intended to remain visible, confirm that the relevant detection corresponds to that person before continuing.
Detection and grouping can be imperfect. Review short appearances, people crossing behind shelves, side views and camera changes. A person absent from the detection list has not thereby disappeared from the footage. Check the exported video and address missed content with an appropriate editing step before disclosure.
Do not assume that all appearances of an individual were grouped together or that the requester remains visible in every needed scene. The review needs to cover the actual result, including the context needed to understand the requester's activity.
Check other identifiers and the selected options
Faces are one part of a CCTV scene. A till receipt, staff badge, number plate, conversation or distinctive uniform can reveal information about someone else. Decide which of these details need attention for the intended recipient.
Video jobs offer options for plates, on-screen text and audio as well as faces. Check the settings for the uploaded file; plate processing is optional. A setting being available does not establish that it was selected, or that automatic detection found every relevant item. Inspect the areas and audio passages you intended to change.
Keep useful timestamps or camera labels where appropriate for the disclosure, while checking that other details do not expose unnecessary information. If an edit makes the requested activity impossible to understand, return to the disclosure decision and adjust the extract rather than assuming that more redaction is always better.
Inspect the file that will leave the organisation
Download the processed copy and play it through at its final resolution. Step through difficult transitions and listen to the audio. Check that the requester and relevant activity remain understandable where that was your intention, and that bystanders selected for redaction are covered throughout their appearances.
There is no claim here that zero customers can be identified from the result. A person may still be recognisable from context even when a face is covered. Where an unresolved detail matters, correct it or seek the responsible reviewer's decision before releasing the file.
Keep a record of the disclosure
Log the source cameras, time range, export version, selected settings and review date. Record who approved the release and any limits or additional edits. Use a distinct filename for the approved copy so it is not confused with the recorder's original export when assembling the response.
Handle retention and delivery through your organisation's access-request procedure. Do not assume a universal recorder retention period or deadline from this page; those questions belong in the applicable procedure and legal assessment.
Start a CCTV video job, check the available options and pricing, and review the exported copy before adding it to the disclosure bundle.
When you need this
A shop is preparing footage relevant to an access request. The operator identifies the cameras and intervals, reviews detected people, chooses redactions and checks the result before assembling the response.
The compliance angle
Access to CCTV must be considered alongside the rights of other people in the footage. The linked EDPB guidance informs that assessment; an edited file alone does not determine the lawful disclosure.
What you can verify
This scenario has no published detection benchmark or guarantee of zero re-identification. For a real request, keep a dated review of the relevant cameras, export version, selected options, remaining identifiers and disclosure decision.
Frequently asked questions
The face review lets you choose among detected identities. Confirm which detection belongs to the requester and inspect the export. It does not guarantee that every person was detected or grouped correctly.