The European Union is moving closer to full enforcement of its artificial intelligence rules under the EU AI Act. From August 2, 2026, new transparency and compliance duties will apply across member states. These rules focus on disclosure when AI is used, labeling of synthetic content, and oversight of general-purpose AI systems.
At the same time, legal uncertainty continues as proposed delays under the Digital Omnibus package remain unfinalized. Penalties for violations may reach up to 7% of global annual turnover in serious cases.
From August 2, providers of AI systems that interact directly with people must make that interaction clear, unless it is already obvious. Providers of systems that create synthetic audio, image, video, or text must also mark outputs in a machine-readable format where technically possible.
Deployers face duties as well. Users of emotion recognition or biometric categorisation systems must inform people exposed to those tools. Deployers of deepfake systems must disclose when image, audio, or video content has been artificially made or changed.
The European Commission says Article 50 transparency obligations apply from August 2, 2026. It has also published a voluntary code of practice for marking and labelling AI-generated content. The legal duties still apply under the AI Act.
The penalty structure gives the August deadline weight. Breaches of banned AI practices can draw fines of up to 35 million euros or 7% of global annual turnover. Other violations, including Article 50 transparency duties, can reach 15 million euros or 3% of global annual turnover.
False, incomplete or misleading information to authorities can bring fines of up to 7.5 million euros or 1% of global annual turnover. Member states must set enforcement rules, and penalties must account for small businesses and startups.
Meanwhile, the high-risk timetable has created a second layer of uncertainty. The Council of the EU and Parliament reached a political agreement on May 7 to delay stand-alone high-risk AI rules to December 2, 2027. Systems embedded in products would move to August 2, 2028.
Even so, companies are treating the change as ‘not yet final’ until the legal process is completed. The original August 2 date still matters for planning, especially for firms using AI in biometrics, hiring, education, infrastructure, or public services.
Italy has moved faster than many EU peers. Its Law 132/2025 created a national AI framework aligned with the EU AI Act. Reports said the law assigned oversight roles to the Agency for Digital Italy and the National Cybersecurity Agency.
On June 10, 2026, Italy’s government gave preliminary approval to two implementing decrees under that law. The draft measures bar fully automated hiring, dismissal, and disciplinary decisions where workers’ rights are affected. They need parliamentary review before adoption.
Italy also shows why compliance work goes beyond legal forms. ISTAT data cited in recent reports showed AI use among Italian firms with at least 10 employees rose to 16.4% in 2025 from 8.2% in 2024 and 5.0% in 2023.
Yet skills remain a blocker. Reuters reported that 58.6% of Italian firms not using AI cited a lack of skills, while 47.3% cited unclear regulation. That gap puts pressure on businesses to map AI tools, classify risks, assign responsibility, and document human oversight before August 2.
Also Read: AI in Search Raises Concerns as EU Probe Into Google Intensifies