Germany's media regulator is in an uncomfortable position: it's supposed to enforce the EU AI Regulation but apparently doesn't know if it's allowed to. The reason is a classic German regulatory problem – and RTL just exposed it.
Two weeks ago, RTL showed five special episodes of the judge show "Ulrich Wetzel – Das Strafgericht" in which Constantin Entertainment and RTL had used AI to modify existing episodes so thoroughly that entirely new stories emerged. The AI use was mentioned in the credits on RTL+ – but not in the linear TV broadcast. This may violate the EU AI Regulation, which has been in effect since August and requires labeling of AI-generated content.
The essentials
- RTL broadcast AI-modified content without labeling on linear TV; mentioned only in RTL+ credits
- The EU AI Regulation (effective August 2026) mandates labeling, for fictional works "in an appropriate manner"
- Media authorities disagree: some (like Lower Saxony's media authority) claim jurisdiction, others don't
- Germany's AI Market Surveillance and Innovation Promotion Act leaves the question unresolved
The rulebook has gaps
The EU AI Regulation is already in force. It requires that content created with AI be labeled – with exceptions for fictional and creative works. But even there, disclosure must occur "in an appropriate manner."
Germany's AI Market Surveillance and Innovation Promotion Act assigns oversight to authorities "responsible under state law." In the media sector, these are the state media authorities. But here's the problem: it's unclear whether these bodies are formally and legally responsible for monitoring compliance with EU labeling requirements – or whether that still needs to be clarified.
RTL argues "appropriate manner" is met
RTL believes it's in the clear. The broadcaster emphasizes that "careful handling of AI Act transparency obligations" is "very important" to them. In journalism and news, the highest standards apply – but in fictional and scripted content, the situation is different:
"New technologies have always been used here; their deployment is established practice. In this context, we must also consider what the AI Act means by labeling 'in an appropriate manner.'"
RTL's argument: the credit on RTL+ fulfills the obligation. Lower Saxony's media authority disagrees, saying labeling in the TV broadcast itself would be "logically consistent."
Industry and regulators are talking – but too late?
RTL says it's "in dialogue with industry and regulators to find practical and uniform solutions for this new, globally relevant topic." That sounds collaborative – but also admits no clear rules exist yet.
The problem: the EU regulation is already live. While broadcasters, regulators, and industry debate, companies operate in legal gray zones. It's also unclear what happens if authorities can't agree – or if one says it has no jurisdiction.
What this means for you
If you work in media, production, or marketing: this gap will close – the question is just how strict. Until then, assume that AI use in content must be labeled, even if the exact form remains unclear. Better to document and be transparent than face explanations later. At the same time, this reveals a fundamental problem: German authorities often take time to implement EU rules – and until then, uncertainty reigns.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
