Freedom of speech/expression does not protect defamation or similar misrepresentations of fact that harm people in any country I know of. Perjury, filing a false report, defamation, confidence-jobs, libel/slander/whatever, etc.
Releasing faked content that imply that what's being presented is true events? Stuff that's clearly harmful to a person? That's already illegal everywhere, but also functionally impossible to actually enforce. Whack-a-mole lawsuits for every viral faked video of Mr Beast or Elon Musk will never work, unless the US gov't creates something equivalent to the DMCA (laws defining how content hosts should police what they host) for this kind of actionable disinformation.
I imagine that in the US, works that are obvious satire or otherwise explicitly marked as satire will, as always, be protected by the Supreme Court. However, videos that simply use AI to fake events about real people? That will stand up, I think. IANAL.
But it doesn't matter. We've already seen that the laws on defamation don't actually work if the disinfo is diffused instead of having a clear identifiable target with deep pockets.
If Fox News posts a faked video of Greta Thunberg explaining her nefarious scheme to destroy the economy, I expect them to get sued again. But if such a video goes viral and ends up on every FWD from Grandma and every conservative facebook group and *chan? Nobody will face consequences.
Stuff like that is already happening *now*, even without AI. Trivial stuff like faking newspaper articles and then posting them as a screenshot gets passed around on X and FB and various messaging tools. Look how many people 100% confidently believe the US election was stolen, or climate change is fake, or vaccines are a nefarious scheme for mind-control, or whatever.
As a thought experiment: if I were to use AI to generate a video of a famous person describing how much they love green beans, and how they’ll do anything to get their hands on a can of delicious green beans, and I post it publicly have I committed a crime?
You used his image to promote green beans without the right to do so. Normally that's something you pay for, you didn't, and I guess he may ask for what the personality rights for such an ad would normally cost. And depending on how you done it, it could conceivably damage to his reputation, qualifying as defamation. I don't know, maybe he got in trouble with the red bean industry because of that.
Assuming there's no actual deeper political or financial implications, no. The standard for libel and slander for a public figure are
1. The content is inaccurate.
2. You knew it was inaccurate (you have to have known they don't feel that way about green beans).
3. The content does measurable harm to their reputation or business.
If that feels like a borderline impossible standard to reach, well, now you know why celebrity tabloids are so prevalent, despite their wealthy, powerful, influential targets hating them. We take free speech very seriously in the United States.
I think the part of this that will be interesting, and that the legal system will have difficulty addressing, is what “inaccurate” means.
At any point in history, I could have written an article about why I believe that a prominent public figure worships the devil and is engaged in a plot to blow up the moon. As long as I didn’t attempt to back up my story with undisclosed facts (“I was allowed access to his email inbox, and now it’s clear to me that he’s trying to blow up the moon”) or lie about a hard fact (“he told me that he is going to blow up the moon”), I would be fine.
The gap for faked media is that you can blur the line between what I am saying about someone and what it looks like they’re saying themselves. So if I make a fake video that appears to be Obama cackling while he describes his plan to blow up the moon, and post it to my YouTube account… am I presenting it as if Obama said these things? Obama doesn’t generally publish content on random tech dudes’ YouTube channels. There’s a line somewhere where the presumption is that the content is faked, but I don’t know of any firm answers so far from the legal system on where they plan to draw that line.
Releasing faked content that imply that what's being presented is true events? Stuff that's clearly harmful to a person? That's already illegal everywhere, but also functionally impossible to actually enforce. Whack-a-mole lawsuits for every viral faked video of Mr Beast or Elon Musk will never work, unless the US gov't creates something equivalent to the DMCA (laws defining how content hosts should police what they host) for this kind of actionable disinformation.
I imagine that in the US, works that are obvious satire or otherwise explicitly marked as satire will, as always, be protected by the Supreme Court. However, videos that simply use AI to fake events about real people? That will stand up, I think. IANAL.
But it doesn't matter. We've already seen that the laws on defamation don't actually work if the disinfo is diffused instead of having a clear identifiable target with deep pockets.
If Fox News posts a faked video of Greta Thunberg explaining her nefarious scheme to destroy the economy, I expect them to get sued again. But if such a video goes viral and ends up on every FWD from Grandma and every conservative facebook group and *chan? Nobody will face consequences.
Stuff like that is already happening *now*, even without AI. Trivial stuff like faking newspaper articles and then posting them as a screenshot gets passed around on X and FB and various messaging tools. Look how many people 100% confidently believe the US election was stolen, or climate change is fake, or vaccines are a nefarious scheme for mind-control, or whatever.