EU AI Office: Political Agreement to Simplify AI Act and Ban Nudification Apps
The European Commission, Parliament and Council reached a political agreement on the Digital Omnibus package, which simplifies the AI Act and introduces an explicit ban on nudification apps. High-risk AI systems will apply from 02.12.2027, and AI integrated into products from 02.08.2028.
This article was generated using artificial intelligence from primary sources.
The European Commission, the European Parliament and the Council of the European Union have reached a political agreement on the Digital Omnibus package, which simplifies implementation of the AI Act while simultaneously introducing an explicit ban on so-called nudification apps. The agreement was announced on 6 May 2026 and represents a compromise between European institutions on how to balance regulatory clarity with the competitiveness of the AI sector.
The Commission states in its official press release that the agreement will “ease the implementation of the AI Act for European businesses while preserving the benefits for European society, safety and fundamental rights.” This is the most significant adjustment to the framework since the AI Act was adopted.
What are the new application deadlines?
High-risk AI systems have an application deadline of 2 December 2027. This category covers biometric identification systems, systems managing critical infrastructure, AI tools in education and employment processes, and systems used in migration, asylum and border control.
The second group comprises AI systems integrated into physical products such as lifts or children’s toys — these have a deadline of 2 August 2028. The difference in deadlines reflects the complexity of certification chains in the product industry compared to software AI services.
What does the ban on nudification apps mean?
The Digital Omnibus introduces an explicit ban on applications that generate non-consensual intimate images of individuals, known as nudification tools. These apps typically use generative AI models to manipulate existing photographs to depict a person without clothing, without their consent.
The ban fits within the EU’s broader framework for protecting dignity, privacy and the rights of victims of digital violence. This is the first explicit pan-European regulatory intervention against this specific category of generative AI.
How does the Digital Omnibus differ from the original AI Act?
The package does not alter the fundamental structure of the AI Act, but simplifies the timeline and calibrates the regulatory burden. The agreement gives industry more time to adapt and the Commission space to develop technical standards and supporting tools before obligations become binding.
The agreement now proceeds through formal adoption in the Parliament and Council, which is the final procedural step before publication in the Official Journal of the EU. More detailed implementation guidelines are expected in the coming months.
Frequently Asked Questions
- When do the new rules for high-risk AI take effect?
- High-risk AI systems (biometrics, critical infrastructure, education, employment, migration, asylum, border control) apply from 2 December 2027, while AI integrated into products (lifts, toys) has a deadline of 2 August 2028.
- What are nudification apps and why are they being banned?
- Nudification apps generate non-consensual intimate images of individuals, most often women, by manipulating existing photographs. The EU is banning them to protect the dignity and privacy of citizens.
- What does the Digital Omnibus package deliver?
- The package reduces regulatory burden by providing a clear application timeline, giving space for technical standards and support tools to be developed before obligations become enforceable, while preserving fundamental rights protections.