Filtering
© Onway Oy 2026
1.1.1. AUTHOR and WORK defined in each sector
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025001
Requirements
"
1.1.1. As rights are granted to authors for their works, AUTHOR and WORK shall be defined for various creative sectors.
Benefits
"
Challenges
"
Notes
"
1.1.2. AUTHORS/CONTRIBUTOR ROLES distinguished/standardised
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025002
Requirements
"
1.1.2. AUTHORS and CONTRIBUTORS shall be distinguished; CONTRIBUTOR ROLES should be standardised by creative sector and declared.
Benefits
"
accurate remuneration of some usages based on industry practices, e.g., standardised royalty splits.
Challenges
"
contributor roles should also be standardised and translated across countries.
Notes
"
1.1.3. AUTHORS/CONTRIBUTORS identified with persistent Ids
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025003
Requirements
"
1.1.3. AUTHORS and CONTRIBUTORS shall be identified with standard and persistent identifiers at the highest possible levels of trust and interoperability.
Benefits
"
Challenges
"
handling anonymity, aliases, and data privacy; dealing with natural and legal persons.
Notes
"
1.1.4. AGENT IDENTIFIERS issued by Trusted Authorities
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025004
Requirements
"
1.1.4. AGENT IDENTIFIERS for AUTHORS and CONTRIBUTORS shall be issued by trusted authorities that recognise each other.
Benefits
"
Streamlined online transactions by exchanging trusted identifiers instead of lengthy descriptors prone to typos and misspellings. Identifiers impose minimum requirements for metadata to be collected about an AGENT, thus ensuring better interoperability. Trusted identifiers can be used to distinguish AGENTS without using any sensitive personal data. Facilitation of authority and access controls. AGENT IDENTIFIERS recorded at the time of creation or publication can be maintained throughout the lifecycle of the digital asset and included in a persistent way.
Challenges
"
AGENT IDENTIFIERS, e.g., ISNI, require a publication or WORK, which imposes timing requirements on the process of binding AGENT and WORK Identifiers as early as possible. It should be possible to acquire an AGENT IDENTIFIER based on preliminary metadata (pre-bibliographic record).
Notes
"
1.1.5. AGENT IDENTIFIERS clearly defined roles
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025005
Requirements
"
1.1.5. Various AGENT IDENTIFIERS shall have clearly defined roles with explicit mappings between them that define when they can be used interchangeably and when they differ in scope or usage.
Benefits
"
Challenges
"
Some AGENT IDENTIFIERS, such as ISNI, are designed to disambiguate public identities, and are intended to be used as bridge identifiers between other databases if a connection to the actual identities of the authors and related information is needed. This is a challenge to interoperability within larger infrastructure.
Notes
"
1.1.6. AUTHORS/CONTRIBUTORS and IDENTIFIERS Resolvable
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025006
Requirements
"
1.1.6. AUTHORS and all CONTRIBUTORS and their IDENTIFIERS shall be unambiguously resolvable for humans and machines within and outside the system identifying them.-- If an opt-out expression is later changed, it is necessary to know who has the right to make such a change. If there are terms and conditions within the rights expression that allow certain uses, it is necessary to know with whom any amendments on these uses should be negotiated. If remuneration models move toward the possibilities outlined in Section 5 of the report, then there will be an increasing need to identify contributors in an unambiguously resolvable manner.
Benefits
"
Challenges
"
Unambiguous resolution of a public identity is a different task from resolving the author’s identity.
Notes
"
1.1.7. Attributions include all CONTRIBUTORS’ names, identifiers…
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025007
Requirements
"
1.1.7. Technically derivable*15 from Art. 19 §4 of the CDSM directive EU/2019/790: attributions should include CONTRIBUTORS’ names, identifiers, roles, and entitlements [CL].
Benefits
"
Challenges
"
Percentage of contribution is not a value that can be expressed in the metadata formats in use in the library sector. To define authors and contributors, which is the main purpose of the library metadata, this information is not necessary or possible to determine in practice. The percentages can be defined in the contracts and agreements between the creators of the work and the publishers, but this is not usually seen as part of or appropriate for the metadata. Percentage of contribution is not set by directives, but by agreements between rightsholders or by CMO practices. However, Article 19.3 states that the obligation set out in paragraph 1 shall be proportionate and effective in ensuring a high level of transparency in every sector. Member States may provide that in duly justified cases where the administrative burden resulting from the obligation set out in paragraph 1 would become disproportionate in the light of the revenues generated by the exploitation of the work or performance, the obligation is limited to the types and level of information that can reasonably be expected in such cases. As such the issue depends on the copyright law of each EU member state.
Notes
"
15 Technically derivable requirements are implicit requirements which should be verified as resulting standards could contribute to a better functioning of the copyright eco-system.
1.1.8. A WORK and its EXPRESSIONS distinguished
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025008
Requirements
"
1.1.8. A WORK (i.e., ABSTRACTION) and its EXPRESSIONS, MANIFESTATIONS, and ITEMS shall be distinguished; EXPRESSIONS, MANIFESTATIONS, and ITEMS shall be unambiguously linked to the WORK they relate to at the highest possible levels of trust and interoperability.
Benefits
"
Challenges
"
Tracking file compressions, e.g., higher resolution for printing and lower resolution for display on the web. Management of trust indicators to achieve the highest possible level of trust. Definition of interoperable identifiers for WORKS, EXPRESSIONS, MANIFESTATIONS, and ITEMS.
Notes
"
1.1.9. WORK IDENTIFIERS issued by Trusted Authorities
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025009
Requirements
"
1.1.9. WORK IDENTIFIERS shall be issued by trusted authorities that recognise each other.
Benefits
"
Challenges
"
Notes
"
1.1.10. WORK IDENTIFIERS defined roles
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025010
Requirements
"
1.1.10. Various WORK IDENTIFIERS shall have clearly defined roles with explicit mappings between them that define when they can be used interchangeably and when they differ in scope or usage.
Benefits
"
Challenges
"
Notes
"
1.1.11. MEDIA ASSET types standardised
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025011
Requirements
"
1.1.11. Types of MEDIA ASSETS should be standardised by creative sector and expressed in metadata.
Benefits
"
Challenges
"
Notes
"
1.1.12. Minimum Set Of METADATA Defined for creation phase
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025012
Requirements
"
1.1.12. A minimum set of METADATA may be defined for specific creative sectors and METADATA exchanges.
Benefits
"
Challenges
"
Notes
"
1.2.1. GenAI output shall be tagged
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025013
Requirements
"
1.2.1. Technically derived from Art. 50 §2 of the AI Act EU/2024/1689*16: applying to any and all Generative AI (GenAI) systems generating any type of media, including generation and manipulation, while not applying to assistance for standard editing: GenAI output shall be tagged when required, and the tag shall be machine readable and detectable. This tagging process should be effective (tagging what should be tagged), interoperable, robust, and reliable.
Benefits
"
Challenges
"
The metadata formats currently in use, for example in the library sector, do not support indication that generative AI has been used. Besides content, Generative AI may also be used in the future to create metadata, which would require additional tagging mechanisms.
Notes
"
16 Codes of practice may be enacted (as long as they are not, they remain recommendations).
1.2.2. GenAI output shall be disclosed
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025014
Requirements
"
1.2.2. Technically derived from Art. 50 §4 of the AI Act EU/2024/1689*17: applying to generation and manipulation of deep fakes and text which is published with the purpose of informing the public on matters of public interest, while not applying when GenAI content is reviewed and/ or edited by human(s) and published under editorial responsibility of a natural or legal person: GenAI output shall be disclosed as such and the disclosure shall be human readable and detectable. Human-made and AI-generated content shall be distinguished and identified.
Benefits
"
Consumers can use metadata to filter materials depending on whether they are generated by AI or not. Encourages responsible use of AI by requiring disclosure.
Challenges
"
Identification of hybrid assets, e.g., a book translated by AI but edited by a human. Definition of minor and major uses of AI. Definition of where human creativity begins and ends.
Notes
"
17 Codes of practice may be enacted (as long as they are not, they remain recommendations).
1.2.3. Tags clear, distinguishable, and immediate
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025015
Requirements
"
1.2.3. Technically derivable from Art. 50 §5 of the AI Act EU/2024/1689*18: tags (from Art. 50 §2) and disclosures (from Art. 50 §4) shall be clear, distinguishable, and immediate.
Benefits
"
Challenges
"
Notes
"
18 Codes of practice may be enacted (as long as they are not, they remain recommendations). confidentiality of intellectual property and sensitive data, or it might be unclear when and to what degree to disclose AI use. When AI is integrated continuously into new tools and software, it may not be clear to the discloser which algorithm has been used. We discussed whether in these cases it would be possible to identify the software used: this may also be problematic if software has various AI-generated functionalities.
1.2.4. AI Algorithms Generate METADATA
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025016
Requirements
"
1.2.4. AI algorithms shall generate METADATA at each step of the training and usage processes of AI systems.
Benefits
"
Challenges
"
Notes
"
1.2.5. AI Algorithms Identified
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025017
Requirements
"
1.2.5. AI algorithms shall be identified.
Benefits
"
Challenges
"
Some creators or organisations may resist disclosing the use of AI algorithms due to competing requirements regarding the
Notes
"
1.2.6. Devices generate METADATA at creation
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025018
Requirements
"
1.2.6. Devices such as text/writing software, cameras, mixing tables, etc. should generate METADATA at the moment of creation.
Benefits
"
Challenges
"
In order to enable easy integration of contributor identifiers into devices and applications used to produce content, a standardised way of including identifiers and other metadata is needed. In this case, devices – and potentially model, firmware version, serial number – should be identified.
Notes
"
1.2.7. Modifications Of A DIGITAL ASSET Identified
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025019
Requirements
"
1.2.7. Modifications of a DIGITAL ASSET shall be identified; multiple modifications of a DIGITAL ASSET shall be tracked and traced, including multiple digitisations of the same DIGITAL ASSET. Technology considerations:-- Hashing may be used. Identification methods such as the ISCC may be used to track various versions of a digital file. Digital signatures and timestamps may be used to increase information trustworthiness.
Benefits
"
Challenges
"
Notes
"
1.2.8. METADATA sets to signal the human-AI distinction
Phase: 1. Media asset creation |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025020
Requirements
"
1.2.8. Standard generic and sectoral METADATA sets shall be expanded with new terms to signal the human-AI distinction.
Benefits
"
Challenges
"
Notes
"
2.1.1. Digital Asset Management Systems (DAMS) Identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025021
Requirements
"
2.1.1. Digital Asset Management Systems (DAMS) shall be identified.
Benefits
"
Challenges
"
interconnections between systems.
Notes
"
2.1.2. Type Of DAMS Identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025022
Requirements
"
2.1.2. Type of DAMS shall be identified, e.g., legal deposit, voluntary copyright registration, recordation of transfers, collective management, etc.
Benefits
"
Challenges
"
Notes
"
2.1.3. CREATIVE WORK distinguished and identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025023
Requirements
"
2.1.3. CREATIVE WORK, original FIXATION, aka master file or master record, EXPRESSIONS, MANIFESTATIONS, and ITEMS of the CREATIVE WORK shall be distinguished and identified.
Benefits
"
Challenges
"
Not all derived works involve new authorship, e.g., a photograph of a painting, a scan of an image, a non-reframed smaller version of a DIGITAL ASSET. The relationship between EXPRESSION, MANIFESTATION, and “version” should be clarified and standardised. Technology considerations:-- In a Digital Asset Management System there are usually two files – an Archive file (master file or preservation file, high resolution) and a Distribution file (potentially lower resolution), e.g., both have hash values in the National Library of Latvia. Indeed, one item may have multiple images connected to it, not only an Archive file and a Distribution file, and connecting their identifications may be required.
Notes
"
2.1.4. OUT-OF-COMMERCE WORKS/ORPHAN WORKS identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025024
Requirements
"
2.1.4. Copyright-protected WORKS, OUT-OF-COMMERCE WORKS, and ORPHAN WORKS and other related subject matters shall be distinguished and identified.
Benefits
"
Challenges
"
Notes
"
2.1.5. FIXATION/time of FIXATION identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025025
Requirements
"
2.1.5. FIXATION, and time of FIXATION – a legal concept – shall be identified at the highest possible levels of trust and interoperability.
Benefits
"
Challenges
"
Notes
"
2.1.6. FILE FORMAT of DIGITAL ASSET identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025026
Requirements
"
2.1.6. The FILE FORMAT of the DIGITAL ASSET shall be identified.
Benefits
"
Challenges
"
Notes
"
2.1.7. DATA FORMAT of METADATA identified
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025027
Requirements
"
2.1.7. The DATA FORMAT of METADATA related to a DIGITAL ASSET shall be identified.
Benefits
"
Challenges
"
Notes
"
2.1.8. A Minimum METADATA set for defined for asset phase
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025028
Requirements
"
2.1.8. A minimum set of METADATA may be defined for specific creative sectors and METADATA exchanges. 2.2. Traceability and transparency
Benefits
"
Challenges
"
Notes
"
2.2.1. Distinguish between parts DIGITAL ASSETS
Phase: 2. Digital asset management |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025029
Requirements
"
2.2.1. Machines should be able to distinguish between parts of a DIGITAL ASSET that are created by different authors and different AI systems.
Benefits
"
Challenges
"
Notes
"
3.1.1. RMI removed or altered by duly authorised people only
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025030
Requirements
"
3.1.1. Technically derivable from Art. 7 §1 of the Infosoc directive 2001/29/EC: RIGHTS-MANAGEMENT INFORMATION shall only be removed or altered by duly authorised people wherever the information may be stored, embedded in WORKS or other subject-matter, or stored in an external database. Therefore, the authority of a person removing or altering RIGHTS-MANAGEMENT INFORMATION shall be verified, hence also the authority of a person adding or entering RIGHTS-MANAGEMENT INFORMATION. Accordingly, the information assertion should be signed and timestamped.
Benefits
"
Challenges
"
Only the authors or other actors with authorisation are allowed to edit (create, read, update, and delete information) rights-management information, referring here to terms and conditions, licenses, and opt-out expressions, etc. If these are edited, who had the right to make the changes must be known.
Notes
"
3.1.2. TRUST FRAMEWORKS designed, developed and deployed
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025031
Requirements
"
3.1.2. TRUST FRAMEWORKS shall be designed, developed, and deployed. They may refer to TRUST RECORDS, TRUST MANIFESTS, TRUST INDICATORS, TRUST PROFILES, and TRUST REPORTS. For example:-- signatures and timestamps should be qualified by listed QUALIFIED TRUST SERVICE PROVIDERS, and attribution and T&Cs assertions may be issued as QUALIFIED ELECTRONIC ASSERTIONS OF ATTRIBUTES.
Benefits
"
Challenges
"
managing different levels of trust, e.g., presence of no timestamp, of a timestamp, of a qualified timestamp. It must be noted that QUALIFIED TRUST SERVICE PROVIDERS and QUALIFIED ELECTRONIC ASSERTIONS OF ATTRIBUTES are mechanisms that have not yet been linked to management of publication metadata and copyright infrastructure, the possible integrations and connections of which are still undefined.
Notes
"
3.1.3. Managers or publishers need mandates to add/alter RMI
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025032
Requirements
"
3.1.3. Most attributions and terms and conditions are asserted by proxy. It is therefore necessary to be able to verify that these intermediaries, e.g., managers or publishers, are mandated to do so.
Benefits
"
Challenges
"
Notes
"
3.1.4. Verification of granularity of mandates
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025033
Requirements
"
3.1.4. It should be possible to verify the granularity of such mandates.
Benefits
"
Challenges
"
Notes
"
3.1.5. Identities, T&Cs and stakeholders must be linked
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025034
Requirements
"
3.1.5. Technically derivable from Art. 7 §2 of the Infosoc directive EC/2001/29: legally-protected RIGHTS-MANAGEMENT Information includes any information, numbers, and code provided by rightsholders to:-- identify AUTHORS or any other rightsholders [NL, CL, IR]; identify WORKS and other subject matters [NL, CL, IR], or assert terms and conditions of use of WORKS or subject matters (T&Cs) [CL]. It is therefore necessary to: identify WORKS (and other subject matters), T&Cs, and stakeholders, and link WORKS and T&Cs, WORKS and stakeholders, and T&Cs and stakeholders.
Benefits
"
Challenges
"
Notes
"
3.1.6. Attribution and T&Cs assertions should be automated
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025035
Requirements
"
3.1.6. Attribution and T&Cs assertions should be automated to some extent. The automated (“by machine”) origin of METADATA should be signalled. Automated METADATA generation reviewed by a human should be considered manual / made by human.
Benefits
"
Challenges
"
Notes
"
3.1.7. GUIs and APIs intuitive, informative, and adaptive
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025036
Requirements
"
3.1.7. Graphic user interfaces (GUIs) and application programming interfaces (APIs) should be intuitive, informative, and adaptive.
Benefits
"
Challenges
"
Notes
"
3.1.8. First fixation/Author’s death shall be recorded
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025037
Requirements
"
3.1.8. As copyright protection is limited in time, the dates of first fixation and the author’s death shall be recorded.
Benefits
"
Challenges
"
Legally speaking, when a person dies, their death certificate records the date of death. But if the body cannot be found, or the person disappears, then the law provides for a waiting period (sometimes up to 7-10 years) before a person can be declared legally dead. This is dependent upon the law of the EU member state; there is no EU-wide standard.
Notes
"
3.1.9. Attribution/assertions valid until superseded
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025038
Requirements
"
3.1.9. Attribution and T&Cs assertions should be valid as long as they are not superseded. Governance, audit, and enforcement processes should be put in place to manage the liabilities of operators of repositories of rights management information. Asset provenance and metadata provenance shall be transparent.
Benefits
"
Challenges
"
Notes
"
3.1.10. Usage policies shall be coherent
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025039
Requirements
"
3.1.10. Usage policies shall be coherent, e.g., an allowed use shall not be in conflict with other permissions and licenses. In the case of earlier agreements on uses with permissions, these must be noted somehow.
Benefits
"
Challenges
"
Notes
"
3.1.11. Misattributions detectable and identified
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025040
Requirements
"
3.1.11. Misattributions of works should be detectable and identified. The resolution of attribution conflicts should be automated – at least partially with human review.
Benefits
"
Challenges
"
Notes
"
3.1.12. Usage definitions for moral and economic rights standardised
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025041
Requirements
"
3.1.12. Usage definitions should be related to both moral and economic rights and be standardised by creative sector.
Benefits
"
Challenges
"
Notes
"
3.1.13. Objectives, rights users’ affiliation, purposes defined
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025042
Requirements
"
3.1.13. Technically derivable from Art. 3 §1 of the CDSM directive EU/2019/790: text and data mining is authorised for reproductions and extractions (objectives) by research organisations and cultural heritage institutions (rights users) for the purposes of scientific research (purposes). Therefore, objectives, rights users’ affiliation (e.g., ‘research organisation’ or ‘cultural heritage institution’), and purposes shall be identified. This would allow “Distribution and consumption” without remuneration to the author when the conditions of the exception under the three-step test apply.
Benefits
"
Challenges
"
Notes
"
3.1.14. Uses superseding three-step test identified
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025043
Requirements
"
3.1.14. Technically derivable from Art. 4 §1 of the CDSM directive EU/2019/790: objectives (reproduction and/or extraction) and purpose (text and data mining) shall be identified (to authorise “Distribution and consumption”) without the need to pay remuneration in accordance with CDSM Art. 19 §1. Uses that are not in line with the three-step test (under CDSM Art. 7) require a license from the rightsholders (opt-in) and must comply with CDSM Art. 19 §1).
Benefits
"
Challenges
"
Notes
"
3.1.15. Rights expression discoverable and interpretable
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025044
Requirements
"
3.1.15. Technically derivable from Art. 4 §2 of the CDSM directive EU/2019/790: Rightsholders may reserve the rights to text and data mining conducted for any other purpose than scientific research. The reservation should be expressed by machine-readable means in the case of content made lawfully, publicly available online. Therefore, the rights expression shall be discoverable and interpretable.
Benefits
"
Challenges
"
Would the absence of a TDM-related reservation in some asset metadata be sufficient to conclude that there is no opt-out, even if such an opt-out reservation has been expressed elsewhere? This is entirely dependent on details of how opt-out declarations are implemented and required to be set up*19: it must be straightforward for AI developers to check opt-out status. When AI-related reservations are very nuanced, whether they are actually usable and actionable by machines may become a challenge.
Notes
"
19 See for example the Code of Practice for GPAI model developers at https://digital-strategy.ec.europa.eu/en/policies/contents-code-gpai
3.1.16. Current standards to include opt-out expressions
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025045
Requirements
"
3.1.16. Generic and sectoral METADATA standards shall be expanded with new terms to signal opt-out expressions and other AI-related expressions.
Benefits
"
Challenges
"
Notes
"
3.1.17. Opt-out expressions both human and machine-readable
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025046
Requirements
"
3.1.17. Opt-out expressions and other AI-related expressions should be human-readable and machine-readable.
Benefits
"
Challenges
"
Notes
"
3.1.18. RMI presented clear understandable format
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025047
Requirements
"
3.1.18. AI-related rights expressions should be granular and relate to specific AI usages such as text and data mining, inferencing, training, training for generation, etc. Nuanced machine-readable expressions shall consider new technical developments.
Benefits
"
Challenges
"
technology is moving fast, so keeping up to date with different AI usages can be a challenge. Indeed, it is currently unknown what will be relevant in the future, or how different AI usages will be conceptualised and named.
Notes
"
3.1.19. Expressions scalable for mixed human machine content
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025048
Requirements
"
3.1.19. Due to the highly iterative nature of AI usage, expressions of AI-related rights shall be highly scalable, for example in the case of human-machine hybrid content (e.g., content produced by several human creators – authors and contributors – and several machines), and in the case of multiple generations of hybrid human-machine mixed contents.
Benefits
"
Challenges
"
Notes
"
3.1.20. Link between assertions robust
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025049
Requirements
"
3.1.20. The link between any manifestation of a particular WORK and METADATA pertaining to that particular WORK – especially assertions of AI-related rights expressions – shall be robust and standardised, especially in cases of dynamic assertions and amendments of the expressions.
Benefits
"
Challenges
"
Notes
"
3.1.21. Assertions not to hinder access to content
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025050
Requirements
"
3.1.21. Assertions of AI-related rights expressions shall not hinder the accessibility of contents as intended by their authors and rightsholders.
Benefits
"
Challenges
"
Notes
"
3.1.22. Coordination across platforms
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025051
Requirements
"
3.1.22. There should be coordination across platforms to ensure that METADATA is unified, coherent, and consistent, including terms and conditions, especially AI-related reservations and rights declarations for particular WORKS. 3.2. Traceability and transparency of metadata management phase
Benefits
"
Challenges
"
Notes
"
3.2.1. Recognise which parts of a work are human/machine made
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025052
Requirements
"
3.2.1. It shall be possible to recognise what parts of a creative WORK have been produced by one or several humans and what parts by one or several machines.
Benefits
"
Challenges
"
Notes
"
3.2.2. Identify membership of a collective management organisation.
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025053
Requirements
"
3.2.2. In some cases, e.g., the remuneration of musicians, it should be possible to identify whether they are members of a collective management organisation.
Benefits
"
Challenges
"
Notes
"
3.2.3. Identify ownership/operation of a generative algorithm
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025054
Requirements
"
3.2.3. It should be possible to identify what organisation owns or operates a generative algorithm.
Benefits
"
Challenges
"
Notes
"
3.2.4. Express and maintain provenance information of DIGITAL ASSETS
Phase: 3. Metadata management |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025055
Requirements
"
3.2.4. The provenance and authenticity of DIGITAL ASSETS should be expressed. This information shall be updateable – e.g., through signed and timestamped claims, and the infrastructure should be able to fetch the updated information reasonably easily.
Benefits
"
Challenges
"
Notes
"
4.1.1. Fair and proportionate remuneration relies on current RMI
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025056
Requirements
"
4.1.1. Technically derivable from Art. 18 (right to fair and proportionate remuneration) of the CDSM directive EU/2019/790: RIGHTS MANAGEMENT INFORMATION shall be reliable, current, and exhaustive, as well as accessible and interoperable.
Benefits
"
Challenges
"
Notes
"
4.1.2. RMI embedded/ linked to all digital assets and remain intact
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025057
Requirements
"
4.1.2. RIGHTS MANAGEMENT INFORMATION shall be either embedded in or linked to all distributed DIGITAL ASSETS and remain intact throughout the distribution process.
Benefits
"
Challenges
"
Notes
"
4.1.3. Licencees need technical capability to track and report
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025058
Requirements
"
4.1.3. Technically derivable from Art. 19 §1 of the CDSM directive EU/2019/790: transparency requires licensees to have the technical capability to track and report:---- exploitation of works and performances; modes of exploitation, to be able to report on the use; revenues generated by online platforms and other licensees, and remuneration due to authors and other rightsholders.
Benefits
"
Challenges
"
Notes
"
4.1.4. Distributors respect and monitor AI/TDM related reservations
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025059
Requirements
"
4.1.4. Distributors shall respect crawler-exclusion tools (robots.txt and the like) and other opt-out mechanisms, and monitor AI-related traffic when so required by the rightsholders of the distributed content.
Benefits
"
Challenges
"
Notes
"
4.1.5. Distributors coordinate respect and standard AI/TDM reservations
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025060
Requirements
"
4.1.5. Distributors should coordinate across platforms to ensure that crawler-related procedures and other opt-out mechanisms support a unified, coherent, and consistent TDM*20 position, and that TDM reservations are expressed in a standardised way.
Benefits
"
Challenges
"
Notes
"
20 TDM = text and data mining. he overall goal of the report. This section encourages open discussion about the requirements that can be foreseen, in this case with commercial licensors, concerning the requirements in their respective sectors.
4.1.6. Validation of user’s eligibility for exceptions
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025061
Requirements
"
4.1.6. Exceptions and limitations require validation of the user’s eligibility. The legislation often includes conditions that must be met in order for the exception or limitation to apply. There is indeed a need to understand for what purpose a copyright-relevant action is taken. Verifying usage may also require verification of users, unless the exception is mandatory for specific purposes.
Benefits
"
Challenges
"
The conditions are based on directives that allow for national flexibilities. Therefore, they may differ within the EU and require specific attention in due course. The data processing of exceptions and limitations shall respect the GDPR and trade secrets. Additional considerations: The following requirements concern primarily the commercial sector and fall outside the scope of National Libraries. The requirements are uncertain, as the infrastructure and legislation are still developing. Nevertheless, they support the overall goal of the report. This section encourages open discussion about the requirements that can be foreseen, in this case with commercial licensors, concerning the requirements in their respective sectors.
Notes
"
4.1.7. Track and report modes of exploitation, sublicensees
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025062
Requirements
"
4.1.7. Technically derivable from Art. 19 §2 of the CDSM directive EU/2019/790 in the case of commercial licensing: transparency requires licensees and sub-licensees to have the technical capability to track and report:---- exploitation of works and performances; modes of exploitation, e.g., archiving, communication to / consultation by the public; revenues generated, and remuneration due. Transparency also requires the capability to identify and report sub-licensees, e.g., customers of an image marketplace. Therefore, rights metadata shall be embedded in or linked*21 to all media assets made available to the public, including attribution and terms and conditions such as use limitations and expiration dates.
Benefits
"
Providing rights metadata helps distributors stay compliant with copyright laws and licensing agreements. It reduces the risk of unauthorised use or distribution, avoiding legal liability. License verification and providing metadata streamlines the management of digital collections, especially across multiple intermediaries and media asset types.
Challenges
"
Some media assets (e.g., older ones, particularly outside National Libraries) may lack standardised metadata formats, making it difficult to consistently embed and read rights metadata. Ensuring compatibility across different platforms, especially with older systems, can be problematic.
Notes
"
21 E.g., through hard binding or soft binding; see Terms and Definitions.
4.1.8. Licencees communicate clearly licence conditions to users
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025063
Requirements
"
4.1.8. When required to align with the scope of the license, digital service providers should plan and enforce access control. Digital service providers shall clearly communicate licence conditions, allowed uses, and usage restrictions at the point of communication to the public. Consumer interfaces should signal the source of the content, its licence, and any attribution or citation requirements. Digital service providers shall provide a reason for denial when content cannot be streamed (e.g., due to licence expiry or regional restriction).
Benefits
"
Challenges
"
Ensuring the accuracy, timeliness, and granularity of licence data is difficult, and metadata upkeep is resource-demanding. Sharing data with rightsholders may raise privacy and legal concerns, particularly under data protection laws such as the GDPR.
Notes
"
4.1.9. Proportionality requires access to standards and technologies
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025064
Requirements
"
4.1.9. Technically derivable from Art. 19 §3 of the CDSM directive EU/2019/790: proportionate obligation requires the identification and description of standards and technologies which allow fulfilment of obligations.
Benefits
"
Challenges
"
the administrative burden should remain proportionate and economically justifiable. Consumer interfaces should reflect any rights expressions, including exceptions, limitations, and reservations; these limitations and reservations shall be easily understood.
Notes
"
4.1.10. Additional remuneration need technical capability to track/trace
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025065
Requirements
"
4.1.10. Technically derivable from Art. 20 of the CDSM directive EU/2019/790: additional remuneration requires the technical capability to track and report usages and related revenues.
Benefits
"
Challenges
"
Notes
"
4.1.11. Automated licensing need standardized types of users
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025066
Requirements
"
4.1.11. Rights expressions, a fortiori in the case of automated commercial licensing, should be supported by the standardisation of: types/identifiers of users to manage standard use cases, exceptions and limitations, e.g., research organisations, cultural heritage institutions, and others; types of usages, referring to rights as defined in the Berne Convention, WIPO Treaty on Copyright, EU Acquis Communautaire, MS legislations, and industry royalty split practices, e.g., different treatment of performing rights splits among Collective Management Organisations.
Benefits
"
Challenges
"
Notes
"
4.1.12. Automated licensing need machine-readable usage policies
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025067
Requirements
"
4.1.12. Commercial licensing, a fortiori automated commercial licensing, should be facilitated by the availability of machine-readable: usage policies, i.e., there is a need for policy engines and intelligent wizards which consider regulations and general industry practices to help generate policies; price structures, and usage contexts, i.e., there is a need for context engines and intelligent wizards which consider regulations and general industry practices to help generate contexts.
Benefits
"
Challenges
"
Notes
"
4.1.13. Awareness and education to authors/performers on licensing
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025068
Requirements
"
4.1.13. In the case of commercial licensing, awareness and education campaigns should be made available to authors, creators, and performers to help them know their rights and benefit from them.
Benefits
"
Challenges
"
Notes
"
4.1.14. Alert services to authors for non-exploitation of works
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: RMI requirement
Reference #2025069
Requirements
"
4.1.14. In the case of commercial licensing, services may be set up to identify and alert authors, creators, and performers who do not seem to exploit their rights. 4.2. Traceability and transparency
Benefits
"
Challenges
"
Notes
"
4.2.1. Provenance assertions embedded/linked and remain intact
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025070
Requirements
"
4.2.1. Going forward, provenance and authenticity assertions should be either embedded in or linked to distributed DIGITAL ASSETS and remain intact throughout the distribution process.
Benefits
"
Challenges
"
Notes
"
4.2.2. Distribution systems shall enable audit trails for governance
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025071
Requirements
"
4.2.2. Distribution systems shall enable audit trails for both internal and external governance. Audit trails shall include claim signatures and timestamps.
Benefits
"
Challenges
"
Notes
"
4.2.3. Downstream usage for AI training traceable and transparent
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025072
Requirements
"
4.2.3. Downstream usage of licensed original content by AI algorithms should be traceable and transparent.
Benefits
"
Challenges
"
Notes
"
4.2.4. DSP log monetised access to content to facilitate remuneration
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025073
Requirements
"
4.2.4. In the case of commercial licensing, digital service providers should log monetised access to content, including asset identifier and other necessary information, to facilitate the proportional and transparent remuneration of authors, creators, and performers.
Benefits
"
Challenges
"
Notes
"
4.2.5. Users shall be informed when and why their activity is logged
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025074
Requirements
"
4.2.5. Users shall be informed when and why their activity is logged, with reference to the digital service provider’s data and rights policies.
Benefits
"
Challenges
"
Notes
"
4.2.6. Logs on usage stored securely and their access controlled
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025075
Requirements
"
4.2.6. Logs shall be stored securely, and their access shall be controlled.
Benefits
"
Challenges
"
Notes
"
4.2.7. End-users view provenance information before consuming
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025076
Requirements
"
4.2.7. RIGHTS MANAGEMENT INFORMATION such as provenance and authenticity assertions should be visible to end-users before they consume the DIGITAL ASSET.
Benefits
"
Challenges
"
Notes
"
4.2.8. End-users view opt-out declarations/right expressions
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025077
Requirements
"
4.2.8. Consumers should be able to view opt-out declarations and any associated right expressions.
Benefits
"
Challenges
"
Notes
"
4.2.9. RMI presented to end-users in clear and understandable format
Phase: 4. Distribution and use |
Date: 2025-01-01 |
Class: T&T requirement
Reference #2025078
Requirements
"
4.2.9. RIGHTS MANAGEMENT INFORMATION shall be presented to end-users in a clear, concise, and understandable format.
Benefits
"
Challenges
"
Notes
"