[AHatLikeThat]

>While it is hard to quantify the prevalence of the AI-generated images, Alenichev and others say their use is on the rise, driven by concerns over consent and cost.

Although I agree that using AI-generated poverty porn is repulsive on many levels, I think there is a valid consent and privacy issue if using actual photos of real people in these circumstances.

A picture of a real person, especially a child who is being victimized by poverty and/or abuse is now being victimized twice.

In this case, AI could be used judiciously to alter the persons face/body and other identifiers to protect their identity, in the same way that print journalism changes names and other identifiers in text. I think this would be ethical as long as properly marked as having been altered and why.

Of course this does not address the motives of the company behind the ad, but it would at least provide increased transparency and privacy protections.