The European Commission confirms unchanged deadlines for AI Act compliance: August 2025 for general-purpose models and August 2026 for high-risk systems. New support measures include the AI Act Serve Desk and Digital Simplification Omnibus package to help companies meet requirements.
As tech companies scramble to prepare for Europe’s sweeping AI regulations, the European Commission has rejected calls for deadline extensions while rolling out new compliance assistance programs. The June 18 launch of the AI Act Serve Desk comes alongside revelations that 68% of firms lack adequate resources to meet the approaching regulatory milestones.
Regulatory timeline remains fixed
The European Commission confirmed on June 20 that implementation deadlines for the AI Act will proceed as originally scheduled: August 2025 for general-purpose AI models like ChatGPT and August 2026 for high-risk applications in sectors including healthcare and critical infrastructure. This decision comes despite mounting industry concerns about compliance readiness.
“We recognize the challenges but believe these timelines strike the right balance between innovation and protection,” stated Commissioner Thierry Breton in a press conference at Berlaymont. The announcement referenced recent data from DIGITALEUROPE showing that more than two-thirds of surveyed companies report being underprepared.
New support mechanisms launched
To address compliance hurdles, the Commission activated its AI Act Serve Desk on June 18 – a real-time guidance system accessible through the Digital Europe Programme portal. Simultaneously, officials proposed the Digital Simplification Omnibus package which aims to reduce documentation requirements by approximately 40%, particularly benefiting small and medium enterprises.
Dr. Linda van den Berg of the Erasmus Centre for Digital Governance commented: “These measures demonstrate pragmatic adaptation while maintaining regulatory pressure. The Serve Desk could prove crucial given the technical complexity of conformity assessments.”
National approaches diverge
France and Germany have taken contrasting paths to implementation testing. France’s €500 million investment in national testing infrastructure contrasts with Germany’s regulatory sandbox approach launched June 20, where companies can test high-risk systems under supervision.
This fragmentation risk was noted by Prof. Henrik Nielsen at Copenhagen Business School: “While sandboxes accelerate real-world validation, divergent national approaches may create uneven playing fields within the single market.” Industry analysts suggest companies may strategically locate development projects based on jurisdictional advantages.
Historical context of tech regulation
The current situation echoes previous technology adoption curves in Europe. When GDPR implementation began in 2018, similar concerns emerged about compliance costs and operational impacts – particularly among smaller firms. However, post-implementation analysis showed most organizations adapted within two years despite initial struggles.
The phased approach of the AI Act mirrors successful elements of earlier EU digital policies like the Digital Services Act package. By staggering requirements based on risk levels rather than applying uniform deadlines across all applications, regulators aim to balance innovation capacity with necessary safeguards – a strategy previously employed during financial sector digitization reforms.
