Image: Olha Solodenko / Shutterstock.com
For decades, copyright law was designed for a fairly simple world: someone writes a text, takes a photo, or composes a song—and no one is allowed to simply copy it. That logic is now beginning to crumble. The reason is artificial intelligence. Modern AI systems don’t learn from scratch. They are fed vast amounts of text, images, music, and code—content that often belongs to others.
As the FAZ notes in a detailed article, the EU now intends to address this very issue. A new report from the European Parliament’s Committee on Legal Affairs makes it clear: things cannot continue as they have been.
The real problem lies in the engine room
Many companies hail AI as a productivity miracle. Text, images, software, analyses—everything gets done faster. What often gets overlooked is that these systems were, in many cases, trained using copyrighted content—that is, works by journalists, authors, photographers, musicians, or publishers.
And that’s exactly where the controversy begins. After all, creative professionals don’t thrive on polite applause; they rely on their work being used and compensated. When AI models are trained on vast amounts of such content without clear consent or compensation, it strikes many creatives as a business model based on the motto: take first, then make billions.
Brussels wants to establish new rules
The report therefore makes a remarkably clear point: traditional copyright law is no longer sufficient for the world of AI. Instead of struggling to apply old rules to new technology, the EU is considering a supplementary framework.
Three main measures are planned. First, rights holders should be able to more easily specify that their content may not be used for AI training. Second, AI providers should be required to be more transparent about what their models were trained on. Third, it’s about money: anyone who builds advanced AI using third-party content should ensure that creators and media companies are not left empty-handed.
This is a politically sensitive issue. Europe wants to catch up in the field of AI, but at the same time protect its cultural and media sectors. Reconciling these two goals will be costly, complicated, and quite contentious.
Why this affects almost every company
This issue is by no means limited to large AI corporations. Companies that use AI in marketing, content creation, customer service, or software development are also affected. In the future, it will be crucial whether the systems used have been properly trained and who is ultimately liable for any potential problems.
Particularly tricky: Content generated entirely by AI is generally not supposed to be eligible for standard copyright protection. So anyone who thinks they can quickly create exclusive content with just a few prompts might end up coming crashing back down to earth.
The real message is clear: the golden age of unrestrained AI training in Europe could soon be over. And that’s exactly as it should be. Innovation is not a free pass to first help oneself to others’ achievements and then pretend that it’s all just technological progress.
Source: faz.net
