Copyright enforcement in the artificial intelligence environment
- Author(s):
- Parliamentary Assembly
- Origin
- Assembly
debate on 23 April 2026 (17th sitting) (see Doc. 16374, report
of the Committee on Culture, Science, Education and Media, rapporteur:
Mr Mogens Jensen). Text adopted by the Assembly on
23 April 2026 (17th sitting).Provisional version
subject to editorial review.
1. The Parliamentary
Assembly stresses that intellectual property rights are a fundamental
catalyst for innovation and investment across diverse industry sectors.
2. The Assembly values the fact that cultural and creative industries
represent a significant economic force in Europe, with a workforce
numbering in the millions. These industries depend on copyright
law for the protection of the rights and interests of authors and
other rightsholders, as well as for the remuneration of their creative
works and contributions.
3. The Assembly acknowledges that the emergence of the artificial
intelligence (AI) era has given rise to a particularly challenging
set of problems for the creative sector.
4. In order to feed their data-hungry systems, AI companies are
scraping the internet without prior permission and without remunerating
content creators on the basis of legislative provisions that are
neither clear-cut nor fit for purpose.
5. AI training requires the making of copies at different stages:
from the initial web scraping and the creation, online publication
and downloading of datasets, to the actual AI training with those
datasets and the use of the resulting mode. Those copies, if copyright-protected,
are acts of reproduction and would require the authorisation of
the relevant rightsholders unless there are covered by a copyright
exception or limitation.
6. The Assembly notes that current European Union legislation
does not offer a solution to this problem. The so-called “text and
data mining (TDM) exceptions”, which were adopted before the advent
of generative AI, place the onus on copyright holders to opt out
of this exception and do not include any remuneration provisions.
7. Moreover, there are doubts regarding the applicability of
the text and data mining exceptions to generative AI, in particular
since these exceptions must comply with Article 5(5) of Directive
2001/29/EC which states that limitations or exceptions apply only
to “certain special cases which do not conflict with a normal exploitation
of the work or other subject-matter and do not unreasonably prejudice
the legitimate interests of the rightsholder”.
8. The Assembly underlines that such legal environment is particularly
advantageous especially for US and Chinese companies. Without a
level playing field, innovation and competition in Europe will suffer.
In the absence of fairness, existing disparities in wealth and power
will be exacerbated. Unfortunately, the present legal system is
incapable of rectifying market failure, as regulators and smaller
competitors lack the financial resources to match the billion-dollar
legal expenditure of tech giants. Furthermore, judicial proceedings themselves
are inadequate for addressing public goods, such as trustworthy
information and digital infrastructure, or externalities, including
disinformation and environmental damage.
9. In particular, the sustainability of the news media ecosystem
may be at risk due to the immediate, short-lived economic value
of news content. It is important to note that lengthy litigation
is not an effective solution to loss of revenue when platforms use
that content without fair payment. During protracted legal proceedings, platforms
can generate advertising revenue and consolidate their market position,
while publishers forfeit revenue that is irrecoverable in the long
term.
10. Otherwise, the impressive capabilities of generative AI tools
in generating new content give rise to other copyright-related issues.
11. There is an ongoing legal debate about the copyrightability
of works created using AI tools, and about who would hold the rights
resulting thereof. Whereas it seems obvious that an AI tool cannot
be a holder of rights, a case-by-case analysis might be required
to determine whether a work created with the intervention of an
AI tool can have a physical person as an author.
12. In any event, it is important to note that content generated
by AI systems that is based on copyrighted material has the potential
to infringe the rights of reproduction, communication and making
available to the public of copyright holders.
13. GenAI content aiming at misleading people (the so-called deepfakes)
are a subject of widespread concern. Deepfakes are not inherently
harmful and can be used for purposes that are legal, such as parody. However,
if misused, they can be used for disinformation purposes and to
manipulate public opinion in electoral processes, and can violate
personality rights by misusing a person’s image and voice. This
violation of personality rights can be particularly harmful when
it comes to the image of minors.
14. In light of all these considerations, the Assembly calls on
the Council of Europe member States to adopt a regulatory approach
that balances the rights and interests of AI providers and rightsholders
so that innovation is not achieved at the expense of creators, and
to protect citizens and democracy at large against the abuse of
AI tools. In this respect, they should in particular:
14.1 clarify in their national legislation
that copyright exceptions such as the text and data mining exceptions
introduced by the European Union’s Directive (EU) 2019/790 on copyright
and related rights in the Digital Single Market are not applicable
to the training of AI systems;
14.2 sign and ratify the Council of Europe Framework Convention
on Artificial Intelligence and Human Rights, Democracy and the Rule
of Law (CETS No. 225) and adopt or maintain measures to ensure that adequate
transparency and oversight requirements are in place to facilitate
the possibility for parties with legitimate interests, including
copyright holders, to exercise and enforce their intellectual property
rights;
14.3 require that providers of AI systems disclose training
data so that rightsholders can assert their rights and provide substantiation
in a court of law for any unauthorised utilisation of their content;
14.4 introduce a rule in their national legislations whereby
it would be presumed that commercial AI systems have been trained
on copyright-protected material in cases where transparency requirements are
not met;
14.5 introduce, in their national legislation, fair remuneration
rules based on an independent valuation for cases where access to
data is not possible, and support collective licensing schemes in
this regard;
14.6 introduce a compulsory final offer arbitration model that
allows a negotiating party to make a request for binding arbitration
to the relevant national ministry when another of the parties has
broken off negotiations, refused a request for negotiations, or
when negotiations do not seem likely to lead to a result;
14.7 require that content generated by AI systems is disclosed
as such by appropriate labelling that is machine-readable, interoperable,
and easily identified by human beings;
14.8 require that the unauthorised distribution of realistic
digitally generated imitations of personal characteristics is considered
illegal in their national legislation;
14.9 require that performers and artists are protected against
the unauthorised distribution of realistic digitally generated imitations
of their performances or artistic achievements;
14.10 promote media and information literacy, and invest in
media and civic education programmes to uphold critical thinking,
especially with regard to the AI tools and the understanding of
its output.
15. The Assembly calls on AI providers to provide transparency
of data used for AI training, and show openness to dialogue and
good will in negotiations with rightsholders.
