At IRISA, researchers are studying the interpretability and explicability of artificial intelligence algorithms, a topical issue particularly in relation to ethics.
Should algorithms explain the reasoning behind their decisions, for example, so that their fairness can be assessed?
Discover the article by Erwan Le Merrer (Researcher Inria, member of the Wide team) "Remote explainability faces the bouncer problem" published in the scientific journal "Nature".
More information