Redact faces before your company shares a video

Before that town-hall recording or training video leaves your organisation for an external audience, pixelate every face in it. Footage you filmed for the workforce becomes a new processing purpose the moment the internal-comms or employer-brand team wants it on the careers page or inside a customer case study, and the employees, contractors and visitors in frame never agreed to that. Medianonymizer destroys the face pixels frame by frame, in support of your GDPR obligation, and hands you a single anonymised export you can attach to the processing record. You can anonymise a video right now without an account.

Why reusing internal footage is a new processing purpose

You filmed the quarterly all-hands and the onboarding session for the workforce — perhaps a sales kickoff, a diversity panel, an offsite retreat or a product demo as well. That is one lawful purpose. Cutting the highlights into a recruitment reel, an employer-brand teaser or a public case study is a different purpose entirely, aimed at an audience the colleagues in the room never expected. Under the purpose-limitation principle, further processing for a new purpose needs its own basis and, where relevant, a compatibility assessment; "we already had the tape" satisfies neither. The employees who walked past the lens, the contractor presenting a slide, the visitor in the back row — none of them signed a release for the open internet, and in many jurisdictions the works council expects to be consulted before staff appear in external marketing at all.

The clean way through is to make the footage stop being about identifiable people before it is shared. Once each face is destroyed, the clip no longer carries the identifier that links it to a named individual, and the external distribution is no longer processing their personal data.

What Medianonymizer removes from the master file

The tool detects human faces across every frame and pixelates them. That is the scope, stated plainly:

• Faces of everyone in shot — presenters, staff in the audience, contractors, visitors walking through.
• Across the full timeline, not just a chosen thumbnail — a face that reappears in a later scene is pixelated there too.
• Destructively, by overwriting the pixels that made the face recognisable.

It does not read lanyards, redact license plates, strip voices from the audio track, or scrub a name from a slide on screen. It anonymises faces in video. Knowing the exact boundary is what keeps the claim honest.

Covering a face versus destroying it

The export belongs in your Article 30 record

Accountability is not a feeling; it is documentation. You get back one new video file whose faces are gone from every frame — there is no companion stream and no hidden master that a subject or an auditor could reconstruct later. That is what makes it useful evidence: attach the file to your record of processing activities, log the reuse against the original recording in your retention schedule, and anyone reviewing the decision can scrub the timeline themselves and confirm that no identifiable employee left the organisation in that clip. The controller keeps the source internally; only the anonymised version travels to the brand team, the agency or the public site.

Where anonymisation sits in the approval chain

In practice the recording passes through several hands: a videographer captures the session, internal communications edits a cut, the employer-branding lead pushes to publish, and you — accountable to the data protection officer — hold the veto. That handoff is precisely where anonymisation belongs. Send the edited cut through Medianonymizer, return the anonymised export to the branding lead, and archive the untouched rushes internally with a dated note of the new purpose. Where your balancing test or a lightweight impact assessment flags the reuse as sensitive, the pixelated version is the mitigation you record. Keep the raw rushes in the corporate media library behind your single sign-on, and let only the pixelated cut reach the content calendar. Where a sub-processor or an outside agency hosts the reel, a data-sharing agreement and a documented lawful basis belong in the same folder; a lightweight DPIA and, if the ICO ever knocks, a dated redaction log close the loop. If a supervisory authority ever asks, it sees a documented, defensible decision rather than a boardroom of exposed colleagues republished across social channels.

What we do not promise

We are precise about scope, because a vague claim is worse than none. This removes faces, and only faces. It does not run your legitimate-interests assessment, certify the licensing of the background music, clear the caption text on a slide, or approve the choice of channel — those governance steps stay with you and your data protection officer. Nor does it add a watermark, burn in subtitles, transcode the container, or maintain your consent register or your records-of-processing spreadsheet. Pixelating the faces is one concrete control in support of your GDPR obligation, not a compliance stamp for the whole campaign. It also does not touch license plates or the audio.

Anonymise the footage before it leaves the building

Upload the master file, let every face be pixelated frame by frame, confirm the price, and download the single anonymised export. Attach it to your processing record, log the new purpose, and share the clip knowing no recognisable employee went out with it. No account, pay only for what you anonymise.