Artificial intelligence is no longer sitting at the edges of Europe’s media sector. Yet adoption remains uneven, and concerns around intellectual property, data protection, editorial integrity and compliance continue to shape how media organisations use (or choose not to use) these tools.

 And while AI offers genuine opportunities to make media production more efficient, improve access to content and develop new services, the systems behind those opportunities depend on data: what is used, where it comes from, what IP rights are associated with it, and under what conditions it can be used. In creative sectors, where AI training can involve vast datasets containing copyright-protected works, these questions are becoming increasingly difficult to separate from AI adoption itself. They have already reached the courts. In Germany, for example, the Munich Regional Court ruled that AI music company Suno had violated copyright and ordered it to pay damages in a case concerning the use of copyrighted music to train its models.

 Europe has already established an important regulatory foundation. The EU AI Act places AI within a framework designed around the internal market, health, safety and fundamental rights. For general-purpose AI, it also introduces transparency requirements concerning model documentation and the content used for training. But the rules relevant to AI do not begin and end with the AI Act. Copyright questions sit alongside this framework, notably through the international treaties on copyright and intellectual property, and EU legislation such as the Copyright in the Digital Single Market Directive. From a data protection perspective, this regulatory foundation also builds on decades of personal data protection legislation, such as the GDPR, and a robust administrative structure and case law.

But regulation alone cannot make AI trustworthy. Legal obligations do not turn values into reality on their own. They also need to be effectively incorporated into business practices, organisational cultures and day-to-day operations.

 There is also an ethical question. Europe’s approach is concerned not simply with what AI can do, but with the conditions under which it is developed and used. That means keeping human rights, accountability and democratic values in view as AI becomes embedded in infrastructure and services. The wider European debate increasingly connects adequate regulation and ethical safeguards with democracy, freedom and diversity.

 The challenge is also operational. Principles such as transparency, provenance and control only become meaningful when organisations can implement them in real workflows. For media organisations, this means being able to establish where data comes from, what rights are attached to it, how it can be used and what happens when it is shared. For AI companies, being able to train their models with data that is identifiable and traceable, with clear copyright licensing, reduces legal risks.

This is also where legal certainty is necessary. When organisations are unclear about the rules, the conditions for participation or the responsibilities involved, moving from experimentation to adoption becomes harder. Clearer conditions can reduce that uncertainty, and trusted data spaces can help create those conditions by bringing governance, data quality, interoperability and clear terms of use into the same environment. Furthermore, features such as observability and traceability are concrete tools to determine where data comes from, who has the right to use it and under which conditions. They also play an important part in supporting compliance with contracts and regulations, such as the EU AI Act and the GDPR.

The trade-off between innovation and regulation is a fallacy. Safeguards, legal certainty and responsible technology can also boost innovation if correctly incorporated, consistently applied and aligned within business processes. The goal is not to choose between innovation and safeguards. It is to make the safeguards part of how innovation happens.

For Europe’s media sector, trustworthy AI therefore depends on rules, values and practical conditions working together: establishing boundaries, guiding decisions and making responsible AI use not just possible, but also attractive in everyday workflows.