Our recommendations and amendments proposal for the AI Omnibus

Europe’s home appliance sector is built on decades of trust, following rigorous safety standards. As we transition into the AI era, the EU's regulatory framework must empower manufacturers to enhance consumer protection, rather than creating a deterrent to safety. 

The current New Legislative Framework (NLF), ensures that the European household appliances which are placed on the market are already safe and fully compliant products. However, we are currently facing a legislative paradox within the text and interpretation of the  AI Act: if a manufacturer integrates Artificial Intelligence to provide a voluntary, additional layer of safety, they risk being reclassified as "High-Risk". 

This classification triggers a cascade of disproportionate administrative and financial burdens that do not reflect the actual risk. By treating voluntary safety enhancements as a liability, we are effectively penalising European innovation and the manufacturers will simply refrain from using the latest technologies to make safer products. 

Below we propose targeted improvements to the AI Act to clarify definitions surrounding AI systems, safety components, and high-risk classification conditions. To illustrate the practical implications of these definitions, the Annex provides two hypothetical scenarios involving home appliances where AI is used to improve safety. We demonstrate how these beneficial features could be inadvertently misclassified as 'High-risk' under the existing text.

1. Improve the broad definition of “AI system” (Art. 3 (1)) 

2. Need to specify the high-risk classification in Article 6 (1)

3. Need to clarify the definition of the Safety Component (Article 3 (14))

4. Interplay of AI Act with Union harmonisation legislation in Section 1 a of Annex I - Clarify the application of conformity assessment

* Please find attached the full recommendations and proposed amendments to the AI Omnibus.