Culture Feature

"In Event of Moon Disaster": Nixon DeepFake Offers an Eerie Glimpse of a Moon Landing Disaster

A team at MIT teamed up with AI specialists to show us what Richard Nixon's famous contingency speech would have looked like.

Popular conspiracy theories would have you believe that NASA—with the help of director Stanley Kubrick—faked the Apollo 11 moon landing in a soundstage in Los Feliz.

In reality there is abundant evidence to the contrary—from the parabolic arcs of moon dust to orbital pictures and retroreflectors—proving that Neil Armstorng and Buzz Aldrin really did land on the moon. On July 20th, 1969—51 years ago today—they ventured onto the surface and took some pictures, but due to damage sustained in their touch-and-go landing, they were stranded there. They never came back…

To Make a Deepfake: Richard Nixon, a Moon Disaster, and our Information Ecosystem at Risk

Keep Reading Show less

Facebook will no longer tolerate any misinformation hosted on its site–except if it's funny.

The $600 billion company recently announced a new policy banning videos manipulated by AI software, also known as "deepfakes." But the policy is a softball to appease critics who malign Facebook's leniency towards political ads that contain false information. Facebook's head of global policy management, Monika Bickert, announced that they will remove videos from Facebook and Instagram according to the following criteria:

"It has been edited or synthesized — beyond adjustments for clarity or quality — in ways that aren't apparent to an average person and would likely mislead someone into thinking that a subject of the video said words that they did not actually say. And:
"It is the product of artificial intelligence or machine learning that merges, replaces or superimposes content onto a video, making it appear to be authentic."

The first loophole is that videos made with less sophisticated software, deemed "shallow fakes," are still allowed to circulate freely, which does nothing to combat the spread of false information contained within more rudimentary videos. Both deepfakes and shallowfakes experience spikes in popularity and invade the internet around election times; Facebook's new policy is clearly part of their preparation for the tumultuous 2020 presidential election.

One notable deepfake in recent years is of House speaker Nancy Pelosi as she seems to slur her words. Shared by many Trump supporters, including Rudy Giuliani, the video is clearly manipulated with basic editing effects found on any smartphone. Since no AI software was used to create the video, it's still permitted under Facebook's new policy.

Likewise, a manipulated video mocking Mark Zuckerberg himself is still allowed to exist on the platform, because it falls under the policy's second loophole: satire and parody. If the video is made for comedic purposes, then it's exempt from the policy. "This is indeed a step in the right direction ... We need to leave breathing room for satire and parody," said Danielle Citron, law professor and vice president of the Cyber Civil Rights Initiative. However, the policy leaves so much room for interpretation that it's bound to be inconsistent. "I would have liked the policy to have included manipulated deceptive media showing someone acted in ways they didn't," Citron said. "Think about deepfake sex videos."

"Facebook wants you to think the problem is video-editing technology, but the real problem is Facebook's refusal to stop the spread of disinformation," a spokesperson for Pelosi said. Aside from the new policy, Facebook will also use "independent third-party fact-checkers" to review videos. If flagged, videos may be permitted to exist on the platform but with a label clearly identifying them as false and manipulated. Those videos will appear less frequently in users' news feeds (and will be outright rejected if they're ads).

Why not remove all false information? Bickert said, "If we simply removed all manipulated videos flagged by fact-checkers as false, the videos would still be available elsewhere on the internet or social media ecosystem. By leaving them up and labelling them as false, we're providing people with important information and context." Okay. But why not label every piece of deliberate misinformation as "false" or "satire?" Why not clearly demarcate fact from fiction in an effort to re-establish the firmament of truth over lies? Maybe because Trump is a sitting U.S. president who is impeached and courting a war with Iran, Australia is on fire, and we're living in an era of celebrity-endorsed misinformation.


The Boundaries of Consent: From One Direction Fan Fiction to the DeepNude App

Where does consent come into play when the people are real but the sex is fiction?


The DeepNude app turns any picture of a woman into a nude.

HBO's "Euphoria" publicly broadcasted an animated sex act between One Direction bandmates Louis Tomlinson and Harry Styles. Currently, over 11,000 R-rated fan fiction stories about BBC's Sherlock are live on Where does consent come into play when the people are real but the sex is fiction?

The DeepNude App


The DeepNude App sounds like the dystopian wet dream of a horned up teenager who grew up watching nothing but Revenge of the Nerds on repeat.

The software has a simple premise: Users upload pictures of clothed women, and the app makes them naked using artificial neural networks to replace shirts with breasts and pants with vulvas. It doesn't work on men.

DeepNude's creator shut the app down within hours of being exposed by Motherboard, reasoning that "the probability that people will misuse it is too high," as if there would be any possible "legitimate" use or that the intent behind its creation wasn't blatant, unchecked sexism. Naturally, copies sprung up immediately, promising to improve upon the existing technology and also removing watermarks denoting the resulting pictures as fakes.

The dangers inherent to DeepNude seem all too obvious. While the technology is far from perfect and many of the resulting images look wonky, some do look almost real, especially at first glance. Unfortunately, a "real-enough" nude picture is more than enough to potentially ruin someone's career, blackmail them, or otherwise damage their reputation. The fact that the breasts and genitalia in the picture are not the subject's own is irrelevant to the potential damage a DeepNude picture could cause someone.

To be clear, this is an app that could also take pictures of children and strip them naked. It's hard to say where the legality of such a thing might fall––would such pictures be considered child pornography if the actual genitals depicted didn't belong to children?––but DeepNude makes a great case study for how our legal system hasn't caught up to our technological advancements yet. Legality aside, DeepNude is so clearly wrong that one could safely assume that anyone who feels otherwise is, at best, completely morally bankrupt and drowning in sexist ideology.

If we can resolutely agree that creating nude photos of people without their consent is morally wrong, we also agree that consent is necessary for fictional, sexual depictions of real people, at least in certain cases. While the idea of real consent for fictional acts may sound silly at first, amidst our cultural reckoning with the definition and boundaries of consent, let's consider how celebrities' consent to sexualized depictions of themselves intersects with the increased virality and visibility of fictional content. The question is where that line of needing consent falls.

The Case of Louis Tomlinson

louis tomlinson euphoria HBO

On a recent episode of HBO's Euphoria, a high school girl obsesses over a fan fiction scene wherein former One Direction member Harry Styles gives his bandmate Louis Tomlinson a blowjob.

The scene plays out via over-the-top, anime-esque animation. It stops just short of depicting any actual privates, but it does show the two band members 69-ing. It's jarring and honestly pretty funny within the context of the show. But more importantly, it uses Styles' and Tomlinson's real names––and while Styles hasn't publicly commented on his unwitting Euphoria cameo yet, Tomlinson has expressed disdain. "I can categorically say that I was not contacted nor did I approve it," Tomlinson wrote on Twitter.

The scene drew a polarized response. Some viewers enjoyed the realism, relating to the exact same early sexual experience of discovering Harry/Louis (or "Larry" or "Larry Stylinson") "real person fiction" (fan fiction featuring real celebrities having sex) on Tumblr when they were younger. On the other hand, many One Direction fans were outraged over the scene and the disrespect shown to Tomlinson, specifically, who had publicly spoken about how the slash/fic fervor made him incredibly uncomfortable and damaged his friendship with Styles. They viewed Euphoria as guilty of publicizing Tomlinson in a fictional sexual act to which he did not consent and had clearly stated made him uncomfortable on prior occasions.

While some fans have suggested Tomlinson sue HBO, the truth is that Tomlinson wouldn't have a case. HBO's depiction of Tomlinson was a legal form of fair use and would qualify as "transformative" in a court of law, meaning there's a substantial artistic/creative element involved with using Tomlinson's likeness. But while the legal implications of Euphoria's exploitation of Tomlinson are cut-and-dried, the morality of the whole ordeal leaves room for interpretation.

Unlike a realistic DeepNude photo, nobody could possibly mistake Euphoria's "Larry Stylinson" scene as an actual depiction of a real sexual encounter between Styles and Tomlinson. At the same time, using the musicians' real names, as opposed to pseudonyms, draws them into a sexual scenario they didn't consent to being a part of. Somehow, morally, that seems incredibly wrong.

To compare: If a man walked up to a woman and described his sexual fantasy about her in detail, most people would consider that to be, at the very least, sexual harassment. He's allowed to have his own private sexual fantasy about her, of course, but he has no right to drag her into it publicly. So why would the morality around that scenario change just because the target in question was a celebrity, or the genders were swapped? The above scenario seems morally wrong regardless. Now imagine that the man proceeded to broadcast his fantasy on television to millions of viewers.

The Boundaries of Fan Fiction

johnlock Michi Shaw via YouTube

Fan fiction is an incredibly important medium for marginalized communities. On top of being a cross-section for fandom and erotica, fan fiction allows many LGBTQ+ youth to experiment with sexuality using avatars that make them feel safe and comfortable. As such, fan fiction communities are largely a force for positivity.

However, certain fandoms take things too far, as was the case with the "Larry Stylinson" community, which actually attempted to push their homoerotic fiction on the real-life subjects of their affection––again, one of whom has openly expressed extreme discomfort about the subject. But while Larry fans might be an extreme example, even fan fiction based on fictional characters have been known to make the actors behind them feel uncomfortable when the fandom is especially vocal.

For instance, in an interview with The Telegraph, Martin Freeman, who plays John Watson on BBC's Sherlock, bemoaned the incessant homoerotic shipping of the "Johnlock" fandom. "Me and Ben, we have literally never, never played a moment like lovers. We ain't f***ing lovers," he said. While there's certainly an argument to be made that Freeman could just ignore fan fiction, the truth is that fans have shoved their fan fiction wishes in his face on multiple occasions, both on Twitter and during live appearances.

In many ways, mediums like Twitter have broken down the relationships between fans and the talent behind their favorite fandoms. The ease of access to creators and actors, and the willingness of some fans to breach personal boundaries due to feelings of entitlement, means that for many creatives, simply ignoring the elements of fandom that make them uncomfortable isn't an option. As such, fan fiction, while largely positive, enters ethical gray area at the point that it personally involves its subjects.

Assuming the content is exactly the same, is there an ethical difference between homoerotic "Johnlock" fan fiction and homoerotic fan fiction about Martin Freeman and Benedict Cumberbatch? It depends. If the fan fiction is never shared and is intended solely for personal use, both instances are pretty okay. Unlike a DeepNude, which (even if made entirely for personal purposes) could be used as a dangerous weapon for victimization, fan fiction doesn't hold the same potential for damage.

But as soon as fan fiction reaches a wider audience, the ones using the actors' real names become morally questionable as they depict real people in a public sexual scenario that they did not consent to. The version using characters, however, is fine, because those characters are not real and have a degree of separation from real actors. In the same vein, as soon as fan fiction, regardless of which names are used, is directly pushed on an actor, it again becomes morally wrong, as it's no different from the hypothetical man describing his sexual fantasy to a non-consenting woman with whom he wants to have sex.

All things considered, fan fiction doesn't need to require consent to exist in a private capacity, but it should never be okay to involve those people without their consent in any public capacity whatsoever. While HBO's inclusion of the real One Direction boys in Euphoria might be legally fine, it's hard to justify morally; it's tantamount to a man sexually harassing a non-consenting woman with his personal fantasies. Consent should always be first and foremost in matters of sex. That doesn't change just because the sex is fictional.