The EU AI Act has been law for months. Starting today, it starts becoming real. The European Commission announced the launch of expert working groups tasked with the technical implementation work that bridges the Act’s legal text and its practical enforcement, specifically, developing standardization frameworks, defining conformity assessment procedures, and producing guidelines for providers of high-risk AI systems.
This announcement matters for a specific reason: the Act’s obligations for high-risk AI providers are not self-executing. They require technical standards to define what “adequate risk management” looks like, conformity assessment procedures to specify how providers demonstrate compliance, and guidelines that translate broad legal categories into sector-specific requirements. That work is now underway. The working groups’ outputs will become the operational backbone of EU AI Act compliance.
There are two distinct implementation tracks running in parallel. The expert working groups announced today are focused on standardization and conformity assessment, the technical infrastructure. Separately, the AI Office has launched its own working groups specifically for developing a Code of Practice, which will provide more targeted guidance for general-purpose AI model providers. Companies operating in EU markets need to monitor both tracks. The conformity assessment procedures affect high-risk system providers directly. The Code of Practice affects general-purpose AI model developers. These are not the same audience and not the same compliance obligations.
Working group membership and specific initial mandates were not disclosed in the Commission announcement. Details are expected as the groups begin their work. What is confirmed: the focus areas are standardization, conformity assessment, and high-risk AI guidelines, the three technical pillars that the Act’s implementation depends on.
For companies building EU AI Act compliance programs, the working group launch is a timing signal as much as a content signal. The standards and guidelines these groups produce will eventually define what documentation, testing, and assessment your compliance program must deliver. Those outputs don’t exist yet. But the clock on developing them is now running, which means the clock on your compliance program’s preparation window is also running.
What to watch: the working groups’ first published outputs, draft standards, preliminary guidelines, or Code of Practice frameworks. Those documents will be the first concrete indication of what EU AI Act compliance will actually require in practice. The gap between the Act’s effective dates and the availability of implementable standards is the compliance window that matters most for planning purposes. Watch the EC Digital Strategy portal for working group progress updates.
The EU AI Act’s implementation machinery is now operational. Companies that have been waiting for technical clarity before beginning compliance planning are running out of runway to wait.