A single formulation in the EU AI Act could delay its entire enforcement by years. An analysis by the Standard shows this: the regulation, celebrated as the first comprehensive legal framework for AI worldwide, is already showing cracks – not due to bad intentions, but because of a gap in the text itself.
The Essentials
- The EU AI Act was praised as a breakthrough in KI regulation, but implementation already shows weaknesses
- The EU failed to deliver harmonized standards on time and provided no practical guidance companies need
- The Digital Omnibus Regulation of July 8, 2026 forced key AI Act start dates to be postponed
- A critical formulation in the legislation could mean delays of years
The Problem: Standards Missing, Deadlines Shifting
The AI Act was ambitious: it aimed to channel KI deployment into regulated pathways without stifling innovation. But early on, a dilemma emerged. The European Commission failed to deliver the harmonized standards on time that companies need for concrete implementation. Without these technical specifications, much remains vague – who knows exactly what "low-risk" or "high-risk" means?
The result: with the Digital Omnibus Regulation on KI of July 8, 2026, corrections had to be made. Key parts of the AI Act were pushed back in time. But that was only a symptom, not the root cause.
The Critical Sentence: Where Regulation Could Fail
The Standard identifies a formulation in the legal text that calls the entire rulebook into question. The exact sentence is not fully disclosed in the report – but the logic is clear: if a central provision is ambiguous or leaves loopholes, companies and authorities could argue over its interpretation for years. While these debates drag on, practical regulation stalls.
This is not an academic problem. It means concretely: companies don't know which KI systems they can deploy, authorities cannot enforce, and the market remains uncertain.
What This Means for German Companies
German firms already operating under GPAI rules (General Purpose AI) or planning to are directly affected. The AI Act was supposed to create clarity – instead, uncertainty grows. Companies must expect compliance requirements to drag on for years while regulation is clarified. That costs time, money, and planning security.
At the same time, it shows: regulation at European level is complex. Even the best intentions and the most ambitious legal framework can fail if practical details don't add up. German companies should not wait for Brussels to answer all questions – instead, they should proactively document their KI systems, assess them, and design them to be robust against different interpretations of the law.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
