What Is Privacy-Compliant Video Sharing?

Privacy-Compliant Video Sharing – Definition

Privacy-compliant video sharing is the lawful and documented disclosure of video footage or images to third parties in a way that ensures the personal data being shared is adequate, limited to the specific purpose, and protected by appropriate technical and organizational measures. In the context of visual material, this primarily concerns facial images, license plate numbers, and other elements that may indirectly identify a natural person.

The main legal benchmark is the GDPR, in particular Article 5(1)(a), (b), (c) and (f), Article 6, Article 24, Article 25 and Article 32 of Regulation (EU) 2016/679. Also relevant are the European Data Protection Board guidelines on processing personal data through video devices, published as Guidelines 3/2019 and adopted in final form on 29 January 2020, as well as CJEU case law confirming a broad interpretation of personal data. In practice, privacy-compliant video sharing means that, before disclosing any footage, the organization should determine the legal basis, recipient scope, purpose of disclosure, retention period, and whether video redaction is required, including face blurring or license plate blurring.

If the purpose of disclosure can be achieved without revealing a full facial image or vehicle identifiers, redaction is generally required under the principles of data minimization and privacy by design. In this sense, redaction is not merely a technical add-on but a compliance measure. For images and video recordings, this usually means anonymization or visual pseudonymization. In day-to-day operations, this most often involves blurring or masking faces and license plates, and manual redaction for elements that are not automatically detected.

When Video Redaction Is Required Before Sharing

Whether redaction is necessary depends on the purpose of disclosure and on whether the third party actually needs to see identifying data. Simply possessing footage lawfully does not automatically mean it can be shared in full. In video materials, the risk of identification is high because data points work together: a face, a license plate, location context, time, clothing, or a vehicle may jointly identify someone.

In practice, video redaction is usually required in the following situations:

• sharing footage with an external entity for incident analysis where identification of bystanders is not needed,
• providing material to the media, a contractor, auditor, or insurer if the purpose does not require disclosure of full visual identifiers,
• publishing training, evidential, or reference materials within an organization where the scope of personal data can be reduced,
• disclosing footage in response to a request where the material includes third parties unrelated to the purpose of disclosure.

For facial images, the duty to protect them arises not only from the GDPR, but also from image rights and personality rights rules. As a rule, a person’s image should be anonymized unless one of the statutory exceptions applies: the person is publicly known and shown in connection with performing a public function, the image is merely an incidental detail of a whole such as a crowd, landscape, or public event, or the person appears only as an incidental part of the overall scene within the meaning of Article 81 of the Polish Copyright and Related Rights Act.

With respect to license plates, the situation is more complex. In many European countries, practice and interpretation lead to an obligation to blur license plates before sharing material. In Poland, there is still some divergence. On the one hand, guidance from data protection authorities and broader European practice support treating a registration number as information that may lead to identification. On the other hand, some administrative court rulings have found that a license plate does not always constitute personal data in itself. From a compliance perspective, the safer approach is a risk-based one, with license plate blurring applied by default unless the plate is necessary for the purpose of disclosure.

How to Understand Anonymization and Redaction in Video Materials

For images and recordings, it is important to distinguish full anonymization from redaction that merely reduces the risk of identification. Full anonymization under the GDPR requires an irreversible removal of the link between the material and a natural person, taking into account all means reasonably likely to be used for identification. In video footage, achieving this standard is often difficult, because indirect characteristics, the background, or metadata may also reveal identity.

That is why visual redaction is commonly used in operational workflows. It includes automatic detection of faces and license plates, tracking them across frames, and applying a blur or cover mask. This approach reduces exposure of personal data, but it does not always amount to full anonymization of the entire material. Documentation should therefore clearly state whether the result is redacted footage, pseudonymized footage, or anonymized footage.

Technologies Used in Privacy-Compliant Video Sharing

Effective video redaction requires computer vision methods. For faces and license plates, detection models based on deep learning are most commonly used. Deep learning is needed at the stage of creating the AI model, which then detects objects in an image and in subsequent video frames. However, privacy-compliant video sharing depends not only on the model itself, but also on quality control, editing, and auditability.

A typical processing pipeline includes the following stages:

• detection of faces or license plates in video frames,
• tracking, meaning maintaining identification of the same object across frames,
• application of a blur mask or cover mask,
• operator review and manual correction of elements not detected automatically,
• export of the redacted version together with process metadata.

An important technical limitation is that Gallio PRO automatically blurs only faces and license plates. It does not automatically detect company logos, tattoos, name badges, documents, or content visible on monitors. These elements can be blurred manually in the editor. Gallio PRO does not perform real-time anonymization or live video stream anonymization. The software runs in an on-premise model, which supports data control and limits the transfer of materials outside the organization.

How to Document the Lawful Sharing of Video Footage

Redaction alone is not enough. For compliance, an organization also needs a decision trail showing why the material was disclosed, to whom, and in what form. This documentation is important as evidence during audits, supervisory authority inspections, and internal accountability reviews under privacy by design.

The minimum documentation set should include:

• the purpose of disclosure and the category of recipient,
• the legal basis under Article 6 GDPR or another applicable provision,
• an assessment of necessity — whether the full material was required,
• a risk assessment for individuals visible in the material,
• the scope of redaction — faces, license plates, and manually blurred elements,
• file versioning — the source material and the redacted material,
• the date, operator, tool used, and quality control outcome,
• the retention period and the secure transfer method used for the recipient.

A simple disclosure form for the material can be very useful. It should clearly separate the source material from the shared copy. If the organization operates in a high-risk environment, it is also worth recording process quality parameters, such as the number of manual corrections or the percentage of frames checked by an operator.

Key Parameters and Metrics for Privacy-Compliant Video Sharing

In privacy-compliant video sharing, formal legal compliance is not enough — the effectiveness of redaction also matters. Parameters should be measured and recorded, because a single undetected face or license plate may undermine the entire process.

If an organization defines its own acceptance thresholds, these should be documented in the procedure. For compliance, high recall is usually more important than maximum processing speed. In practice, it is better to accept more manual corrections than to risk leaving a face or license plate unblurred.

Normative and Interpretive References

Privacy-compliant video sharing should be based on primary legal and regulatory sources. In the European context, the most important sources are the legal acts and guidelines relating to data protection and image processing.

• Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 — GDPR.
• EDPB, Guidelines 3/2019 on processing of personal data through video devices, final version of 29 January 2020.
• CJEU, case C-212/13 Ryneš, judgment of 11 December 2014 — broad interpretation of processing involving video surveillance.
• Article 29 Working Party, Opinion 05/2014 on Anonymisation Techniques, 10 April 2014 — anonymization limitations and risks.
• ISO/IEC 27001:2022 — information security management system standard, relevant for access control, retention, and transfer.

It is worth noting that interpretive differences mainly concern license plates. For that reason, any decision to disclose them should always be justified by the purpose and a risk assessment. As a precaution, blurring license plates is usually the more legally defensible option.