Facial recognition already has a questionable track record. Which is why I am disappointed (but not surprised) to see similar technology applied to age verification.

From a recent article in Les Echos:

« On est confrontés à une réaction plutôt violente quand on refuse la vente aux mineurs. L'algorithme permet d'avoir un appui. Même si ça n'annihile pas les comportements agressifs, ça coupe souvent court aux discussions », témoigne Gaëlle Veillet, buraliste dans la Savoie,

(Source: Les Echos, "L'IA au service de buralistes pour détecter les mineurs")

Loosely translated, the excerpt notes that one shop owner likes the system because it shuts down arguments. (Patrons won't argue with the machine.)

Let that sink in for a moment. I often point to the Little Britain "Computer Says No" sketch when talking about AI. This is an example of a shop owner actively leaning into that idea because it works in their favor.

I wonder how this same person would feel if they were rejected based on a black-box AI system?