"Technical error" sounds frustratingly vague and contentless, but it's odd to think about the lack of cultural consensus about whether the ability for others to depict someone with generative AI is more a benefit to him or her, or more a risk. I could imagine someone hearing this and sincerely thinking that Meta was protecting Lula from deepfake attacks (making it slightly harder for people to falsely depict him doing saying embarrassing or offensive things). Uses of AI to mock or attack people often get more attention than uses of AI to praise them.
In the U.S., if someone had programmed a generative AI to be unwilling to depict Obama, I imagine many people would assume that decision was meant to help Obama, not harm him.
In the U.S., if someone had programmed a generative AI to be unwilling to depict Obama, I imagine many people would assume that decision was meant to help Obama, not harm him.