It’s bound to happen, BUT this problem isn’t unique to AI images; anyone could make these with photoshop. The problem is how quickly social media, without proper safeguards, can spread fake images AI generated or not.

I’m getting the impression that governments will be quick to put guard rails on AI, but are still dragging their feet on guardrails for social media.

anyone can download and start training a model toward specific figure that have enough exposure to the media. Banned by midjourney won’t stop anything really. (it just shows that midjourney have a lot of celebrity photos in their dataset. )

Please put the article link in the post title instead of the body.

On the article: I’m not too worried about cheating photos, since political discourse really should focus on the politics instead of personal drama. On the other hand, AI could be really destructive if somebody e.g. faked a video on police brutality. Paris is currently rioting after a video leaked of an officer shooting a non-threatening driver, and we aren’t too far off from being able to fake such videos.