TL;DR
Since August 2, 2026, significant changes have occurred in AI regulation, including delayed high-risk obligations and ongoing compliance requirements. This report clarifies what is confirmed, what is still uncertain, and why these developments matter.
Since August 2, 2026, the European Union’s high-risk AI obligations have been deferred, but key transparency and disclosure rules remain in effect, affecting AI providers and users across the continent. This shift impacts compliance planning and regulatory expectations for AI developers and deployers.
On June 29, 2026, the Council of the EU approved the Digital Omnibus on AI, delaying the implementation of high-risk obligations for stand-alone systems until December 2, 2027, and for embedded AI in regulated products until August 2, 2028. Despite the delays, many transparency and disclosure mandates, including chatbot disclosures, machine-readable markings, deepfake labeling, and public-interest AI content disclosures, remain scheduled to take effect within days.
The delays were part of broader negotiations after the AI Act, enacted in August 2024, faced implementation hurdles such as incomplete standards, unestablished competent authorities, and capacity issues. The Omnibus also introduced notable changes, including a new prohibition on AI systems generating non-consensual sexual imagery and a limited GDPR-side allowance for processing sensitive data for bias detection, both effective from December 2, 2026.
While the postponement of high-risk obligations has caused some to believe the regulation is effectively on hold, the core transparency rules and specific obligations like AI-generated content labeling continue to apply, creating ongoing compliance requirements for AI providers operating within the EU.
Implications of Regulatory Delays and Ongoing Obligations
This development underscores the EU’s cautious approach to AI regulation, balancing delays with strict transparency and disclosure mandates. For AI developers and companies, understanding these rules is crucial to avoid penalties and maintain compliance. The persistence of certain obligations, despite delays, emphasizes the EU’s commitment to transparency and accountability in AI deployment.
As an affiliate, we earn on qualifying purchases.
EU AI Regulation Timeline and Negotiation Milestones
The EU AI Act, effective since August 2024, set ambitious deadlines for high-risk AI systems, but by late 2025, implementation faced delays due to incomplete standards and capacity constraints. The final approval of the Digital Omnibus on June 29, 2026, shifted key deadlines, notably deferring high-risk obligations by over a year. Despite this, core transparency and labeling rules were retained, signaling a continued regulatory focus on AI accountability.
Negotiations involved multiple trilogues, with initial disagreements leading to a provisional agreement in May 2026. The final approval in June 2026 was nearly a last-minute compromise, avoiding a scenario where high-risk AI regulation would be enforced without harmonized standards in place. The regulation’s phased implementation remains a work in progress, with some obligations already in effect and others delayed.
“The Digital Omnibus ensures a balanced approach, deferring certain obligations while maintaining essential transparency and safety measures.”
— European Commission spokesperson
Scanlily Smart QR Label System Using AI for Inventory and Organization (90 White 2cm Diameter Stickers)
- AI-Powered Inventory Creation: Easily create item databases with AI assistance
- Free CSV Export: Export inventory data without subscription fees
- Ideal for Business and Moving: Suitable for businesses, moving, and storage needs
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Remaining Uncertainties About Full Regulatory Implementation
It is still unclear how quickly national authorities will establish the necessary standards and capacity for enforcement, especially given the delays. The full impact of the deferred high-risk obligations on AI deployment and compliance strategies remains uncertain, as does the timeline for final delegated acts and standards.
Additionally, the extent to which non-compliance with transparency rules will be enforced or penalized in the coming months is not yet known, leaving some ambiguity for AI providers operating within the EU.
A Practical Guide to Detect GenAI Content
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps in EU AI Regulation Enforcement and Standards Development
The European Commission is expected to publish final delegated acts and standards in the coming months, clarifying compliance requirements. Member states are tasked with establishing national AI sandboxes by August 2027, which will influence enforcement and oversight. Monitoring how authorities implement these obligations will be critical for AI companies aiming to stay compliant.
Further negotiations and updates are anticipated as the EU finalizes its oversight framework, with ongoing discussions around standards, capacity building, and enforcement strategies.
AI-Powered Contract Management: AI-Powered Contract Management:AI contract management, legal automation, contract lifecycle management, AI legal tech, … compliance monitoring, smart contracts.
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What are the key AI transparency obligations that remain in effect?
Providers must disclose when users are interacting with AI systems, ensure machine-readable markings on generated content, and label deepfakes or AI-generated public-interest content, among other requirements, all effective within days of the regulation’s latest amendments.
Are high-risk AI obligations completely postponed?
No. The obligations for high-risk AI systems are deferred until December 2, 2027, but many transparency and specific disclosure rules remain active, requiring ongoing compliance efforts.
What new restrictions were added in the Digital Omnibus?
A new prohibition was introduced against AI systems that generate non-consensual sexual imagery and child sexual abuse material, effective from December 2, 2026. Additionally, a limited GDPR-side allowance for processing sensitive data for bias detection was added, with strict safeguards.
When will the EU finalize standards and enforcement capacity?
The European Commission is expected to publish delegated acts and standards in the coming months, with national authorities establishing AI sandboxes by August 2027. The timeline for full enforcement remains to be seen.
How should AI companies prepare given these delays?
Companies should continue to implement transparency and disclosure measures required now, monitor regulatory updates, and prepare for future compliance obligations as standards and enforcement capacity develop.
Source: ThorstenMeyerAI.com
