This sub-category consists of the following datapoints:
1. A Platform Unique Identifier (PUID)
2. Specifications on the content that is affected by the statement
Why is this information included?
Article 17(1) and 17(4) of Regulation 2022/2065 (DSA) require the statements of reasons submitted to the database to be clear and specific. The requirements under this first sub-category also increase the database’s function to ensure transparency and to enable scrutiny over content moderation decisions in line with recital 66 of the DSA.
This is a string that uniquely identifies this statement within the platform’s systems.
The PUID allows localisation of the exact information that is affected by the decision and the Statement of Reasons within a platform’s own systems. It offers the opportunity to connect the data in this database with the information that was affected by the decision, including the URL where the information is/was hosted by the platform. The characteristics of the PUID are to be decided by the platform. It is a requirement that every PUID is unique and may not be re-used in the future.
The specifications of the content affected by the decision cover the required fields regarding the type of content affected and the content creation date, as well as optional fields concerning the language of the affected information.
This attribute describes the type of content that is subject to the restrictions imposed by the statement of reasons. Content can fall into more than one of the following categories: App, audio, image, product, synthetic media, text and video. For example, a post on a social media platform could consist of text and image content. Similarly, a video modified by (generative) AI tools could be described as both, a video and synthetic media. If none of the before-mentioned options adequately describe the type of content subject to the restrictions imposed by the statement of reasons, “Other” can also be selected.
If the selected option for the attribute content_type was content_type_other, a further specification is required. Please only indicate the nature of a content type that is not included in any of the other categories of the content_type attribute.
This attribute indicates the date of the creation of the content, i.e. when the platform started hosting the information affected by the decision. For example, this can be the date that a user account was registered, or the date that specific content was posted.
This attribute concerns the language of the information affected by the decision. For example, the language selected as the user’s account settings, or the language in which the content was posted.
If ‘other’ was chosen as content_language, this field allows to specify the exact language. Please only indicate a language that is not included in the list provided under content_language.
This sub-category consists of the following datapoints:
3. The type of restrictions imposed
4. The duration of the restriction
5. The territorial scope of the restriction
Why is this information included?
Article 17(3) of Regulation 2022/2065 (DSA) sets out minimum requirements for the information that is to be included in a statement of reasons. This includes information on the type of restriction(s) imposed as well as their territorial scope and duration, as specified in Article 17(3)(a). In accordance with Article 17(1) (a)-(d), the types of restrictions included are visibility restrictions, monetary payment restrictions, service restrictions and account restrictions.
Each statement of reasons needs to include at least one decision regarding a restriction imposed by the platform. One of the four types of restrictions (visibility, monetary, provision of the service, service’s account) mentioned in article 17(1)(a)-(d) must therefore be selected.
However, one decision and statement of reasons can contain multiple restrictions. For example, a post on a social media platform that has been deemed allegedly illegal under a particular national law could be subject to a visibility restriction and lead to a temporary suspension of the recipient of the service’s account. Similarly, a counterfeit product offered on a marketplace could be subject to a visibility restriction as well as a restriction of monetary payments. It is therefore possible to select multiple types of restriction decisions for each statement.
As such, at least one of _decision_visibility, decision_mandatory, decision_provision, decisionaccount is mandatory. In addition, depending on the individual circumstances of each statement of reasons, platforms should select each additional applicable decision type.
This attribute describes the visibility restriction(s) imposed on specific items of information provided by the recipient of the service. The restriction can fall into one or more than one of the following categories: Removal of content, disabling access to content, demotion of content, age restriction of content, interaction restriction of content and labelling of content. If none of these options adequately describe the visibility restriction(s) imposed by the statement of reasons, “Other” can also be selected.
When the decision contains a visibility restriction that is different from one of the specific restrictions included under decision_visibility, then this field is required to indicate the nature of the restriction.
This attribute describes the monetary payment restriction imposed on specific items of information provided by the recipient of the service. The restriction can fall into one of the following categories: Suspension of monetary payments or termination of monetary payments. If none of these options adequately describe the monetary restriction(s) imposed by the statement of reasons, “Other” can also be selected.
When the decision contains a monetary restriction that is different from a monetary suspension or monetary termination restriction, then this field is required to indicate the nature of the restriction.
This attribute describes the restriction imposed on the provision of the service to a recipient of the service. The restriction can fall into one of the following categories: Partial suspension of the provision of the service, total suspension of the provision of the service, partial termination of the provision of the service or total termination of the provision of the service.
This attribute describes the restriction imposed on the recipient of the service’s account. The restriction can fall into one of the following categories: Suspension of the account or termination of the account.
Each restriction decision has a temporal scope determined by its start and end date. If a statement of reasons includes multiple restrictions, the duration can be different for each type of restriction imposed.
For example, a visibility restriction could be imposed indefinitely on a piece of content that was deemed allegedly illegal. Simultaneously, the recipient of the service’s account could be suspended for three months for the same infringement. In this case, the duration of the two restrictions imposed would be different, for they would have different end dates.
Depending on the restrictions submitted under _decision_visibility, decision_mandatory, decision_provision, and decisionaccount, the relevant end date attribute values need to be provided. Leaving end date attributes blank indicates that the relevant restriction was imposed indefinitely.
Taking the example above, the end_date_visibility_restriction attribute would be blank, whereas the end_date_account_restriction attribute would be set to a date three months after the application_date attribute.
This is the date, from which the restriction(s) applies/y.
This is the date, when the (longest – if multiple were provided) imposed visibility restriction ends.
This is the date when the monetary restriction imposed ends.
This is the date when the service restriction ends.
This is the date when the account restriction ends.
Territorial scope (territorial_scope) This is the territorial scope of the restrictions imposed. Each value selected must be the 2 letter ISO code for the respective country and the countries must be EU or EEA countries. Multiple countries can be selected.
If the scope of the decision is the entire European Union (EU) or European Economic Area (EEA), all EU or EEA countries can be selected at once in the webform.
This sub-category consists of the following datapoints:
6. A description of the facts and circumstances
7. Information on the source of the investigation
8. Information on the account affected by the decision
Why is this information included?
Article 17(3) of Regulation 2022/2065 (DSA) sets out minimum requirements for the information that is to be included in a statement of reasons. This includes the facts and circumstances relied on in taking the decision, as specified in Article 17(3)(b).
The facts and circumstances of each content moderation may be different. It is important that, in line with Article 24(5) of the Regulation, online platforms should not include personal data in their submissions. Online platforms should be particularly careful with free text fields in this respect.
Facts and circumstances relied on in taking the decision (decision_facts)
This is a field to describe the facts and circumstances relied on in taking the decision. Platforms are encouraged to use standard language for similar cases.
This attribute describes what led to the investigation of the content, which is part of the facts and circumstances relied on in taking the decision. The source of the investigation can fall into one of the following categories: An investigation can be based on a notice submitted in accordance with Article 16 DSA, a notice submitted by a trusted flagger or any other notification that is not considered a notice under Article 16. In case of notifications of incompatibility with terms and conditions that do not qualify as notices under Article 16, the ‘other notification’ option should be selected. While Article 17 of the DSA does not cover orders under Article 9, orders from authorities that do not fall under Article 9 can be indicated as a reason for investigation under the ‘other notifications’ option. Notices submitted by trusted flaggers are also notices under Article 16, but if they are submitted by a trusted flagger, this option should be selected. Alternatively, an investigation can be based on a platform’s own voluntary initiative.
In accordance with Article 17(3)(b) of the Regulation, the identity of the notifier needs to be included in the statement of reasons, but only if that is strictly necessary to identify the illegality of the content. This can be the case, for example, for infringements of intellectual property rights.
Even in such cases, providers of online platforms shall ensure that the information submitted does not contain personal data, in accordance with Article 24(5) of the DSA.
Account type (account_type)
This attribute concerns the nature of the account connected to the information addressed by the decision. It indicates whether the account holder is a business user in the meaning of Regulation (EU) 2019/1150 or the account was a private account, meaning an account for personal use only.
This sub-category consists of the following datapoints:
9. Information on the use made of automated means for the detection of infringements
10. Information on the use made of automated means for taking decisions on infringements
Why is this information included?
Article 17(3)(c) of Regulation 2022/2065 (DSA) requires the parties that submit statements of reasons to provide information on the use made of automated means in taking the decision, including information on whether the decision was taken in respect of content detected or identified using automated means.
This attribute automated_detection indicates whether and to what extent automated means were used to identify the specific information addressed by the decision. ‘Yes’ means that automated means were used to identify the specific information addressed by the decision.
This attribute automated_decision indicates whether and to what extent automated means were used to decide on the infringing nature of the specific information addressed by the decision. ‘Fully automated’ means that the entire decision-process was carried out without human intervention. This attribute does not refer to the identification of such content, but solely to the decision taken after the identification. ‘Not automated’ means that the entire decision-process was carried out without the use of automated means. ‘Partially automated’ means that both automated means and human interaction were applied in the process of taking a decision regarding the infringing nature of the information.
This sub-category consists of the following datapoints:
12. For allegedly illegal information: the legal ground relied upon
13. For allegedly infringing information: the contractual ground relied upon
14. Overlap between incompatibility and illegality
15. Reference (URL) to the legal or contractual ground
16. Category & Category specification
Why is this information included?
Article 17(3)(d) and (e) of Regulation 2022/2065 (DSA) require the parties that submit statements of reasons to include a reference to the legal ground or the contractual ground relied on when taking the decision. This requirement also includes an explanation as to why the information is considered to be illegal or incompatible with the referenced ground. A category indicating the type of illegality, or the type of TOS incompatibility, based on which the content was moderated, must be selected. The categories allow queries for information necessary to enable scrutiny over content moderation decisions.
This attribute indicates whether the decision was taken in line with article 17(3)(d), meaning that the information was allegedly illegal, or in line with article 17(3)(e), meaning that the information was allegedly incompatible with the service’s terms and conditions.
This is a field where the exact legal ground (i.e. the applicable law(s)) that was/were relied upon in taking the decision must be stated.
This is a field to explain why the information is considered illegal on the basis of the legal ground indicated. The explanation does not have to repeat the facts and circumstances but can refer to those.
This is a field where the exact contractual ground (i.e. the relevant section in the applicable terms and conditions) that was relied upon in taking the decision must be stated.
This is a field to explain as to why the information is considered incompatible with a specific section in the service’s terms and conditions. The explanation does not have to repeat the facts and circumstances but can refer to those.
Some information that is incompatible with the terms and conditions of a service can simultaneously be illegal. This optional field allows to indicate whether information that was restricted on the basis of an alleged incompatibility with the terms and conditions was also considered illegal by the online platform. Leaving this optional field blank does not imply that the online platform considers that the content is not illegal.
Where a specific URL to the legal or contractual ground is available, it is encouraged to include these to allow for a quick identification of the ground that was invoked to take the decision.
Examples are a direct URL to the applicable version of the terms and conditions relied on in taking the decision, or a direct URL to the applicable law relied on to take a decision on the alleged illegality of information. An example of the direct reference to the DSA is: https://eur-lex.europa.eu/eli/reg/2022/2065.
A list of categories and specifications are included to codify the type of illegality and/or the type of TOS incompatibility that led to the restriction of the information. The category and specifications section consists of three attributes:
Firstly, a high-level category classification must be indicated. The high-level classification indicates the main category under which the grounds relied on in a statement of reasons fall. The list of high-level categories is exhaustive, and a single selection is required. It is important to select the option that best covers the grounds based on which the decision was taken.
For cases in which more granularity is required, the additional two – optional – attributes category_addition & category_specification can be used to further clarify the high-level category. The category_addition attribute provides a list of category options identical to the high-level category classification. This list is also exhaustive, but it allows for multiple selection. Lastly, the category_specification attribute allows for the addition of keywords that further specify the high-level categories. Platforms are encouraged to use as many additional category_addtion and category_specification options as necessary to describe the grounds relied on in taking the decision as precise and specific as reasonably possible. The lists have been designed as follows (listed in alphabetical order), where the numbered listings concern the high-level classifications, and the sub-categories concern the category specifications. The category_specification options also contain an option to select “Other”:
Animal welfare
Data protection and privacy violations
Illegal or harmful speech
Intellectual property infringements
Negative effects on civic discourse or elections
Non-consensual behaviour
Online bullying/intimidation
Pornography or sexualized content
Protection of minors
Risk for public security
Scams and/or fraud
Self-harm
Scope of platform service
Unsafe and/or illegal products
Violence
Not captured by any high-level category
For example, if an unsafe product was removed from an online marketplace that was also a counterfeit good, the relevant statement of reasons could indicate either “Unsafe and/or illegal products” or “Intellectual property infringements” as the main category, depending on which one was considered more relevant to the decision by the platform. The respective other option could be chosen in category_addition and one or more of the relevant category_specification options (e.g. “Trademark infringement” and/or “Dangerous toys” or other category_specification options considered relevant by the platform) could be added to provide additional information. Platforms are encouraged to use as many options as they see fit to best describe their statements of reason in category_addition and category_specification.
For those statements of reasons where platforms feel that no existing specification properly reflects the grounds they used, this section allows for a free text submission to add specifications. The Commission will monitor the submissions in this free text field and update the options in category_specification accordingly, should recurring submissions prevail.