Video Anonymization After a Supervisory Authority Order: How to Comply with a UODO Deadline

TL;DR: Once you have received a formal order from the President of UODO concerning photos or recordings, the priority is not to build a long-term compliance programme, but to carry out the order quickly, demonstrably, and within the deadline. You need a repeatable 5-step process: identify the materials, prioritise them by reach, anonymize them, document the work, and confirm completion to the authority. Gallio PRO, as on-premises software, automatically blurs faces and license plates, while other elements can be corrected manually in the editor - locally and without storing logs.

Visual data anonymization means processing a photo or recording in a way that prevents a person from being recognised or a vehicle from being identified based on visible identifiers - primarily faces and license plates. In a crisis scenario, when an organisation has already received a formal order, the DPO, legal team, PR team, and technical team need answers to four questions: which materials must be anonymized, how to prioritise them, how to process files without creating additional risk, and how to document that the order has been carried out diligently. In an urgent scenario, what matters is a repeatable process, not declarations.

What Does a UODO Order Usually Cover in Relation to Photos and Recordings?

If the order concerns the publication of visual materials, organisations are most often required to remove or anonymize elements that make it possible to identify a person. In videos and photographs, the main risk area is the image of a person’s face. License plates are also often involved, especially when the material shows road traffic, a car park, a fleet, or the surroundings of company premises.

It is worth distinguishing two levels of analysis from the outset. The legal level answers the question of whether a given element should be treated as personal data or as a personal interest. The operational level answers the question of what exactly needs to be blurred in order to comply with the order on time. In a crisis, the second level must be as simple and well documented as possible.

The obligation to anonymize faces does not arise automatically from a single provision. It depends on an assessment of the lawfulness of publishing a person’s image and processing data in a specific case - in particular under the GDPR, the Civil Code, and copyright law [1][3][4]. Copyright law provides exceptions to the requirement to obtain consent for disseminating a person’s image, for example when the person is widely known and shown in connection with the performance of public functions, when the person is merely a detail of a larger whole such as an assembly, landscape, or public event, or when dissemination of the image has been agreed in exchange for payment. However, these exceptions should not be used automatically as a defence once the authority has already issued an order. At that point, the safer business practice is to comply with the order within the scope indicated in the decision while organising the legal argumentation in parallel for any further action.

Action Plan After Receiving an Order: 5 Steps to Meet the Deadline

1. Identify All Photos and Recordings Covered by the Publication

The first step is to determine the full scope of the materials. This should include not only source files, but also versions published on the website, in social media, press materials, landing pages, campaign archives, and assets shared with subcontractors. In practice, scattered publication channels are the most common reason for missing the deadline.

At this stage, it is not worth starting mass processing yet. First, create a simple register: file name, publication location, material type, visible faces, visible license plates, other elements requiring manual redaction, and completion status. This creates an evidence base in case of a follow-up inspection.

2. Prioritise by Risk and Publication Reach

The second step is to set priorities. Materials that are still publicly available and widely indexed should be processed first. Next come files used in paid campaigns, the newsroom, social channels, and partner materials. Archives with limited exposure should come last.

In practice, it is useful to apply a simple rule: the easier it is to find the material and the more clearly a face or license plate number is visible, the higher the priority. This approach is often seen as a reasonable risk management method, even if full compliance with the order requires work to be carried out in stages.

3. Process the Materials: Automation for Faces and License Plates, Manual Redaction for the Rest

The third step is the actual visual data anonymization. In an urgent scenario, on-premises video anonymization software works best because it allows the organisation to process its own files locally without spreading the workflow across multiple services. This type of solution is used for anonymizing photos and videos with a focus on faces and license plates.

A very precise distinction is needed here. Gallio PRO’s automatic detection covers only faces and license plates. Gallio PRO does not blur entire body silhouettes, does not perform real-time anonymization, and does not anonymize video streams. It also does not automatically detect company logos, tattoos, name badges, documents, or content displayed on monitor screens. If the authority’s order also covers such elements, manual redaction in the editor is required.

That is why, in a crisis, it is worth splitting the work into two tracks. The first track covers fast face blurring and license plate blurring across the entire collection. The second covers manual redaction of non-standard elements that the system does not detect automatically. When handling a larger number of files, batch work across multiple materials and later quality control can be helpful - you can check the workflow by downloading the free demo.

4. Document Compliance with the Order

The fourth step is just as important as anonymization itself. The organisation should be able to demonstrate which files were analysed, which elements were blurred, who approved the result, and when the material was replaced in publication. It is worth keeping a “before and after” summary, but with access control and a minimised number of authorised users.

The tool’s architecture also matters here. From a risk reduction perspective, the absence of unnecessary logs containing detection data is a strong argument. Gallio PRO does not collect logs containing face and license plate detection data, nor logs containing personal data or special categories of personal data. This can have practical significance when assessing whether the process of complying with the order has created a new problematic dataset.

5. Inform the Authority and Close the Publication Loop

The fifth step is to formally confirm that the order has been carried out. This usually involves a short letter describing the scope of actions, the implementation date, the list of publication channels, and whether some materials were temporarily taken down for further redaction. In communication with the authority, it is better to avoid general statements. Provide the number of files, publication locations, and the anonymization method used.

If the case involves a large media library, local deployment, specific security requirements, or the need to reproduce the workflow across multiple organisational units, it is reasonable to get in touch with the team to discuss the deployment model and work organisation.

Decision Table: What to Blur First After an Order Is Issued

Faces and License Plates: Where Automation Ends and Operator Responsibility Begins

In a crisis, it is easy to make one of two opposite mistakes. The first is assuming that running a tool is enough and the issue is closed. The second is assuming that every second of footage requires manual editing. The sensible approach lies in the middle: automation speeds up work on the most common visual identifiers, but responsibility for the scope of the order remains with the organisation.

That is why a final review is needed after the material has been processed. Especially in dynamic recordings, poor lighting, partial face obstruction, or shots taken from a long distance, operational effectiveness depends on the quality of the material. Any claim about speed or accuracy should be treated as context-dependent unless supported by measurable data for a specific dataset.

How to Approach License Plates When the Case Concerns Poland

License plates require caution. There is no uniform rule stating that blurring them is always mandatory across Europe under EU law. In Poland, the situation is also not entirely clear-cut and depends on the context in which the material is used and the real possibility of identifying a person. Data protection authorities and EU case law emphasise a broad understanding of personal data [2][5], but this does not automatically mean that every license plate in every piece of material is always personal data.

Once an order has been issued, theoretical disputes usually do not help with meeting the deadline. From a business perspective, the safer approach is to apply a precautionary standard and blur license plates if the material remains publicly available or is to be submitted to the authority as evidence of compliance with the order.

Common Mistakes When Complying with a UODO Order on Visual Materials

The first mistake is limiting the work to a single publication channel. The second is anonymizing only faces even though license plates are visible in the material. The third is failing to manually check elements that the system does not detect automatically. The fourth is not keeping a register of completed actions. The fifth is leaving working versions of files in circulation without clear access rules.

In practice, an effective response does not have to be complicated. It must be complete, repeatable, and demonstrable. In an urgent situation, this is exactly the operating model a DPO needs.

FAQ: Video Anonymization at the Request of a Supervisory Authority

After a UODO order is issued, is it better to remove the recording or anonymize it?

It depends on the wording of the order and the technical options available. A common practice is to temporarily remove the material from publication and then republish it after anonymization. When the deadline is very short, this approach reduces current risk.

Is blurring faces always enough?

Not always. If the material contains license plates or other elements that indirectly identify a person, face blurring alone may be insufficient. The scope must be assessed based on the wording of the order and the publication context.

Does Gallio PRO automatically detect all personal data visible in a recording?

No. Automatic detection covers only faces and license plates. Logos, tattoos, name badges, documents, and content displayed on monitor screens require operator assessment and, if necessary, manual redaction in the editor.

Does Gallio PRO work in real time for live broadcasts?

No. Gallio PRO does not perform real-time anonymization or video stream anonymization. This article concerns a post-factum scenario in which the authority has already issued an order relating to existing materials.

Do entire body silhouettes need to be blurred?

There is no such general technical assumption. Gallio PRO does not blur entire body silhouettes; it blurs faces and license plates. If the authority has indicated a broader scope, individual analysis and possible manual redaction of other elements will be required.

Can logs from the anonymization process create additional risk?

Yes, if they contain detection data or other personal information. Gallio PRO does not store logs containing face and license plate detection data, nor personal data or special categories of personal data.

How can you demonstrate to the authority that the order was carried out on time?

Usually through brief, specific documentation: a list of processed files, publication locations, completion date, a description of the anonymization method used, and confirmation that the materials were replaced in public channels. The more measurable the description, the better.

Related Materials

• Video anonymization with Gallio PRO
• Download the free demo
• Documentation of video anonymization for a GDPR auditor: sample report and register
• Manual video anonymization: when automatic detection is not enough

Prepared by the Gallio PRO team - specialists in data protection and video engineering who develop anonymization software used in security, the public sector, and the media. This material is for informational purposes only and does not constitute legal advice.

Received an order and facing a short deadline? Test the video anonymization workflow on your own files with the free Gallio PRO demo.