Why are Deepfakes Still Legal?

A bit of trivia for you this morning.  While I have not finished all of the My Ballots for each year (they're not official until I put them on this blog, I've decided), at this point, all of the live-action Star Wars films have something in common, save for one-they all, no matter how much I like them, have won a nomination for Best Visual Effects.  Whether it's the original Lucas trilogy or the prequels or the newer legacy sequels, even the maligned Solo offshoot, they're all in my top five nominees for a given year.  The only one of the films that is without such a nomination is Rogue One.  This is a weird snub, because honestly Rogue One's visual effects look incredible.  The entire water planet explosion in the last twenty minutes easily earns a nomination on its own.  But the film is one of the first movies to utilize a type of de-aging/deepfake technology that at once feels very uncanny valley (depriving us of a lot of the emotional response you'd get from an actual human being), but also it feels really ethically grey.

De-aging is not new in film, and in most cases, it's not particularly convincing.  It's been done in movies as vast as The Irishman, Pirates of the Caribbean 5, Guardians of the Galaxy 2, and Captain Marvel, sometime convincingly (probably the best example is still how they did it for The Curious Case of Benjamin Button), and other times (like The Irishman) with a sea of floating heads.  But Rogue One stands apart because it's the only major film of this type to do this sort of work with a deceased actor, in this case Peter Cushing (Carrie Fisher was still alive when she was featured in a de-aging situation in the same film).  Despite assurances from the film's creators and the cooperation of his estate (they said that they worked with his family), Cushing never officially approved of being part of this movie, he never had a choice to have this be a part of his legacy.  The clips in this film were not clips from past movies where he had a say in the matter-this was using his likeness without his knowledge, without his approval.  That has always felt really icky to me.  Your family shouldn't get to decide what you do with your body-only you should.

You might be wondering why I'm bringing this up seven years after the movie (particularly since I talked about this a lot when the movie came out).  This is because I was struck by the ethics of deepfake technology and, that we haven't advanced enough beyond it this past week when a Twitter account (I'm not sharing a link to it-google if you want), started to use video clips of Jurassic Park and the TV series The Bear where they altered the appearance of the actors in these properties, specifically their eyes, to have them look directly at the camera.  Aside from the fact that the images, which were meant to be realistic but felt dehumanizing (the original clips were more natural) were ugly, it also felt really wrong.  Without context, it would be pretty easy to assume that these clips were real, and while this wasn't necessarily made with malice, it's easy to see how it could be & it still feels like they're violating these actors rights to their body by changing the looks without their consent, and it begs the question-why is this not illegal?

To date, only three states have made deepfakes illegal: Virginia, Texas, & California, and in all cases the scope of this ban is narrow.  Texas & California's laws focus specifically on deepfakes being used in a political campaign, while Virginia is largely going after deepfakes in pornography.  These laws certainly wouldn't ban something like what the Twitter account did, just changing the look of an existing work of art.  You could try to approach it as violating the copyright laws of the companies that own Jurassic Park and The Bear (in fact, that's probably your best route), but what does that do for non-copyrighted material?  If someone did this to, say, a video they found of you on the internet...what recourse would you have?

A lot of the conversations about deepfake technology, or defending against these laws fall into two camps: one, that people are going to do it anyway, and two, that it violates the Constitution to ban deepfake technology.  The first is a stupid argument, and one I've never entirely understood as justification for any ban (this is frequently something levied against gun control legislation).  People also continue to rob, kidnap, and murder...those are still crimes.  The law means that there's a recourse for those wrongdoings...we need to do the same for banning deepfakes.

The second half is much harder to get around, but it's time we start having this conversation publicly, before deepfakes become the norm & it's harder to do something about it.  The Supreme Court has usually sided with First Amendment advocates when it comes to a person publicly lying.  US vs. Alvarez is a recent example of this, where the Supreme Court stated that the Stolen Valor Act, which banned someone from lying about winning the Medal of Honor, violated the First Amendment.  Basically, telling a lie in itself is protected speech.  While there are aspects of deep fake technology that could fall under existing libel laws, for example if someone put out a deepfake video of you saying profanities without your consent, if they are just doing things like using your likeness without your consent and without profit...does that violate their First Amendment rights?  The Supreme Court has consistently struggled with new technology, particularly when related to the First Amendment, and while it is a good avenue to generally side with the First Amendment all-things-being-equal, I think we're entering an era where the truth is unknowable if we don't start putting some rules around deepfakes, and while it's inevitably going to lead to court discussions, it's time that our legislative bodies started to put rules around what other people can do with your likeness without your consent (or after you've died).