Ethical AI must be accessible, understandable, and useful to everyone, not just those who are already connected, educated, or privileged. Technology cannot become a factor of social exclusion or an advantage reserved for a few. If AI reinforces inequalities instead of reducing them, it fails its purpose.

Digital barriers do exist: in territories, in bank accounts, in education levels, in bodies, in peripheries. They are distances that separate the global North and South, young and old, secure and precarious workers.

An ethical project cannot ignore these imbalances: it must make them visible and help eliminate them.

Accessibility is not just technical (simplified interfaces, voice commands, language localization).

It's also cultural and symbolic: we need ethical literacy, which helps people understand not only how to use AI, but also when it's right to trust it, what to ask for, and when to resist.

Finally, justice means fairness in data; an AI that relies only on information from rich or culturally dominant environments reproduces partial visions of the world. A fair technology is one that can see even those who have no voice in the data.

Accessibility and justice are not options: they are ethical foundations.