Potentially, but I don't think in practice, because most image generation is not trying to be photorealistic. Most models are trying to stylize their outputs, e

Potentially, but I don't think in practice, because most image generation is not trying to be photorealistic. Most models are trying to stylize their outputs, either by default (e.g. a Miyazaki-anime-like model), or in response to the prompt. Using AI image generation for misinformation is by far the vast minority of its use. And if the output is stylized, it's going to be much more obvious if it has come from an AI model, because a human is going to have a much harder time reproducing a specific ML model's style on-demand (e.g. if their art teacher asks them to sketch a face in the same style, to prove their homework wasn't faked).