Faces and plates blur themselves,
before a photo is public, not after a complaint
Photos and video taken in public spaces, at events, in store footage, in street-level real estate shots, often capture faces and license plates. Those should not be published without consent, and checking every frame by hand does not scale. We build an agent that detects and blurs faces, plates and other personal identifiers automatically before content goes out. A human spot check covers anything it flags as uncertain.
Why the stray face in the background is a real problem
Photos and video from events, street-level shoots, store footage or real estate exteriors often capture bystanders’ faces. They can also capture vehicle license plates that were never part of the intended subject. Publishing that content without masking them is a privacy exposure under GDPR and similar laws in many markets. Checking every photo or every frame of video by hand for incidental faces and plates does not scale past a small volume. It is exactly the kind of careful, repetitive check that a tired reviewer under deadline pressure is most likely to miss.
Inconsistency is the next problem. One reviewer blurs conservatively. Another misses a plate in the background of a wide shot. Your actual privacy exposure ends up depending on who happened to review that piece of content.
Turnaround suffers too. Content that needs a privacy pass before publishing either waits for someone with time to review it carefully, or it gets published without the check under time pressure. That is the exact failure mode this automation exists to prevent.
None of this shows up as one dramatic failure. It shows up as a steady drag. Work that should take minutes stretches into a backlog item. The quality bar holds on a quiet week and slips on a busy one. The team knows the fix is mechanical, but it never finds a free afternoon to build it.
What gets blurred automatically
The agent detects faces, license plates and other common personal identifiers in photos and video. It applies a blur or mask automatically before the content reaches a publishing queue. On video, it tracks detections across frames instead of treating each frame on its own. That handles movement and partial occlusion better than a naive frame-by-frame approach.
Detections it is not fully confident about are flagged for a quick human check instead of being left unmasked or masked wrong. That covers a partially obscured face or a plate at an odd angle. A consent list lets your team exclude specific people, staff in branded content, models who signed a release, from automatic masking where that is the right call.
Typical integrations: your content management system or photo and video storage as the source. The destination is your publishing pipeline, whether that is your website, social accounts or marketing materials.
What legal still signs off on
Final compliance judgment on edge cases stays with your legal or compliance team. That includes any decision about what level of masking a specific market or use case actually requires. The agent is a technical control, not a legal opinion. Maintaining the consent list, who has actually agreed to appear unblurred, is a human-managed process. The agent never infers it on its own.
How it earns trust before going live
Every masking decision, including anything a human reviewer corrected, is logged, giving your compliance team an audit trail if a privacy question comes up later. Detection confidence starts conservative. More content gets flagged for human review than will eventually be needed, and the threshold loosens only once accuracy is proven on your actual content type.
Before it runs unattended, we run a side-by-side dry run against a sample of your own content. Your team sees exactly what it would have done. Every build ships with a short written runbook, so your team can pause it, adjust a threshold, or roll it back without waiting on us. The running-cost estimate below is a starting budget you set, with an alert built in before it is crossed.
Price and timeline
| Option | Price | What it covers | Timeline |
|---|---|---|---|
| Single automation | from $600 | Face detection and blurring across photos and video | 5 to 10 days |
| Department package | from $2,500 | privacy masking, image moderation and document recognition across your compliance and content team | 2 to 4 weeks |
Running cost is usually $10 to $80 a month in model usage depending on volume, with a budget cap set before launch.
Related
Pair this with image moderation for marketplaces if your platform also needs content policy checks alongside privacy masking. It also pairs with receipt and ID document recognition for the document side of personal data handling. For the compliance side of broader policy rules, see compliance checklists.
The full package breakdown is on the AI agents service page and the development service page. For a real build involving privacy-sensitive infrastructure, see the secure messenger protocol case study and the visa centre support bots case study.
Ready to stop reviewing every photo for stray faces and plates by hand? Get in touch and we will test it on a real content batch in the first call.
Tired of doing this by hand? We can take the whole routine off your team, not only this step: Routine takeover, from $400 →
FAQ
How much does face blurring and GDPR masking cost?
from $600 to set up detection and masking for your content type, live in 5 to 10 days.
Does it work on video as well as photos?
Yes, detection and masking run frame by frame on video, tracking faces and plates across movement rather than only catching a single still frame.
What if the agent misses a face or blurs the wrong thing?
Detection confidence is tuned conservatively, and anything uncertain is flagged for a human check rather than left unmasked or masked incorrectly without review.
Can we exclude specific people who have given consent to appear unblurred?
Yes, a consent list can be maintained so staff, models or anyone who has explicitly agreed to appear is excluded from automatic masking.
Does this guarantee full legal compliance with GDPR or similar laws?
It is a technical control that reduces risk significantly. Final compliance judgment, especially on edge cases and your specific jurisdiction, stays with your legal or compliance team.