AI Act Delay: A False Sense of Relief for Companies
The European Union is considering delaying key obligations of the AI Act, especially those related to high-risk AI systems, potentially pushing enforcement to 2027 or even 2028.
This shift reflects a growing gap between regulatory ambition and operational reality. While the AI Act sets one of the most comprehensive frameworks globally, focusing on risk classification, transparency, and governance, most organizations are simply not ready.
Mapping AI systems, assessing risk levels, and implementing compliance processes require time, expertise, and tooling that many companies still lack. Meanwhile, AI adoption continues to accelerate, often ahead of governance structures.
The delay offers short-term relief, reducing immediate compliance pressure. But it also introduces uncertainty: should companies slow down and wait for clearer rules, or continue building and risk future rework?
One thing remains unchanged: the direction of travel. Regulatory expectations around accountability, traceability, and transparency are not going away. If anything, market pressure, from regulators, partners, and users, is already forcing companies to act.
The delay is not a pause. It’s a test.
Companies interpreting this as “more time before acting” are making a strategic mistake.
In reality:
- Compliance maturity will become a competitive advantage
- Early movers will lock trust, partnerships, and market access
- Late adopters will face compressed timelines + higher costs
The real risk isn’t non-compliance in 2026.
It’s being unprepared when compliance suddenly becomes a commercial requirement.
The smart move today is not to wait for perfect clarity, but to:
- Start mapping AI use cases
- Identify potential high-risk systems
- Build lightweight governance frameworks
Compliance is no longer just legal—it’s becoming a go-to-market asset.
Source
Solutions Numériques – https://www.solutions-numeriques.com/ai-act-repousse-leurope-temporise-face-a-une-realite-plus-complexe-que-prevu/