A recent arXiv preprint proposes a computational approach to comparing normative ethics theories with the European Union's Artificial Intelligence Act. The study uses transformer-based semantic textual similarity to measure alignment between the Act's text and the language of different ethical frameworks.
The authors note that the EU AI Act is a comprehensive regulatory framework that explicitly invokes fundamental rights and ethical principles. Their method appears designed to assess whether the Act's ethical commitments map onto established normative theories, such as deontology, consequentialism, or virtue ethics.
Because the abstract is truncated, specific findings and the set of theories examined are not fully available. The paper is listed as a revised version (v5), suggesting iterative development. As a single source, it offers no independent comparison; the significance lies in its novel use of semantic similarity to bridge philosophy and regulation.