Libertarian Answer Man: Deepfakes and Revenge Porn

From Facebook:

Libertarian Answer Man time:
Deepfakes, Revenge Porn, etc.

Someone asked me if I had written or had any thoughts on the issue of whether AI and applications like deepfake videos or photos could be a rights violation, for example if it is used to make a fake video of someone depicting them doing things that they actually did not do. This could be done to titillate some audience or even for financial gain, and/or to tarnish someone’s reputation, for revenge, and so on. Apparently some states are considering legislation to ban deep-fake porn, and so on But new laws in California and the demand for new legislation against Deep-fake pornography make me take pause at the unintended consequences this may have on artwork, film, political cartoons and expression itself.

My reply:

I have not written much on it but I think it’s an easy case–it follows directly from the Rothbard-Block perspective on defamation law and my approach to intellectual property rights. Of course it doesn’t violate anyone’s right, *per se*, to make a deepfake. (By the way I don’t think most of the deepfakes are very good, for now at least, and I am skeptical that true AI is even possible, but that is a bit beside the point.)
There should be no copyright, no reputation rights, no personality rights, and so on, so someone making an image or video that resembles your face — doesn’t violate your rights.
The one exception would be if you do this in order to cause harm to someone. For example I have argued that speech, in general, is normally not a rights violation–you argue this even more strongly than me, in opposing even incitement as a crime. (I think incitement can be a crime.) But consider Walter is falsely accused of rape by Anne. Stephan goes on the witness stand and lies: he claims he witnessed Walter rape Anne. So Walter gets convicted and imprisoned. I believe in this case, Stephan’s speech–his testimony–played a causal role in the aggression done to Walter by the legal/penal system. So this is an example when speech is aggression.
Likewise, if I use my AI or computer tech to make a fake video showing Walter raping Anne, and this is used in court to convict Walter, again, this would be an illicit use of a deepfake.
But in general, deepfakes violate no rights since there ought to be no copyright, no name and likeness rights, no reputation rights.
So for example suppose Stephan and Walter have competing pizza restaurants. Stephan makes a deepfake showing Walter peeing in his pizza crust in secret, and posts it on the internet. So Walter loses customers and Stephan gains customers. Stephan has done something immoral but not any crime since he didn’t breach any contract and Walter has no right to his reputation nor to his potential future customers. Stephan has only been dishonest and he lied to people, but that’s not per se a crime.

***

From the comments: