From Ethics to Effective Legal Protection: The Role of Legal Protection by Design in the EU AI Act

Abstract

This paper examines two “by design” approaches relevant to the EU AI Act’s effort to balance the risks and opportunities of artificial intelligence (AI): Ethics by Design (EbD) and Legal Protection by Design (LPbD). EbD seeks to embed ethical principles in AI development but lacks enforceability and mechanisms for closure and balancing, limiting its effectiveness in safeguarding fundamental rights. LPbD, by contrast, incorporates legally binding protections into technology, providing enforceable and systematic safeguards. The paper analyses the interplay between these approaches, showing how LPbD addresses EbD’s shortcomings while strengthening accountability under the AI Act. It concludes that EbD can play a complementary role, but LPbD must remain central to ensuring that trustworthy AI respects fundamental rights.