- Prime Minister Modi Targeted by Deepfake: Cyber Law Expert Raises Concerns and Provides Safety Measures Times of India
- Govt to meet social media platforms to discuss deepfake issue; IT Minister warns immunity will not apply if platforms don’t take action The Tribune India
- Gravitas: Indian PM Modi calls out deepfake videos WION
- Rashmika Mandanna, deepfakes and us: I find myself constantly questioning the safety of uploading personal information and disclosing my travel location The Indian Express
- Inside The Universe Of Cheapfakes, Deepfakes | How To Combat this AI Info-calypse India Today
Tag Archives: deepfake
Fork me: ‘Fall’ movie removed more than 30 F-bombs with deepfake dub technology
Without enough production budget for reshoots, the director of upcoming action-thriller Fall says the team turned to AI technology to remove over thirty F-bombs to turn its R-rating into a much more box office friendly PG-13, Variety reports.
The problem — which has now turned into a handy little marketing hook — apparently emerged when the indie film was picked up by Lionsgate for a cinematic release, where an R-rating (meaning children under the age of 17 cannot see the film without an adult present) would limit its box office potential when it releases in the US on August 12th.
“When we were filming the movie, we didn’t know if we were R or if we were PG-13, so I said the F-word so many times,” one of the film’s stars Virginia Gardner said. “I think [director Scott Mann] wanted to kill me in post when we were trying to get a PG-13 rating.” Thanks to machine learning, the final movie reportedly includes family-friendly lines like: “Now we’re now stuck on this stupid freaking tower in the middle of freaking nowhere.”
Variety reports that the swaps were made possible thanks to the film’s director Scott Mann coincidentally serving as co-CEO of Flawless, a company that specializes in using its TrueSync AI technology to translate films between different languages. Its technology is designed to offer “seamless” lip-sync that make it appear as though the film’s original actors are speaking and performing in an entirely different language.
“For a movie like this, we can’t reshoot it. We’re not a big tentpole… we don’t have the resources, we don’t have the time, more than anything else,” Mann said in an interview. The film was shot with IMAX cameras in the middle of the Mojave Desert in California on a modest production budget of just $3 million, meaning that reshoots would have cost time and money that simply wasn’t available. “What really saved this movie and brought it into a wider audience was technology,” Mann said. Variety reports the virtual redubs were completed in under two weeks.
Although altering a film before its original release generally isn’t as controversial as edits made once it’s already in cinemas (*cough* Maclunkey), it always feels like a shame when a director’s original vision doesn’t get a public release. And a small-ish indie film like Fall seems unlikely to see an uncensored director’s cut released after its initial cinematic run.
While Fall used AI to change individual words, there are hopes that machine learning could allow entire movies to be made available seamlessly in different languages, without the telltale lip-sync issues that make current dubbing efforts such an eyesore.
In 2020, Polish film The Champion became the first film to be entirely virtually redubbed into another language (English), which it did thanks to technology from Tel Aviv-based startup Adapt Entertainment. VFX-focused YouTube channel Corridor Crew did a breakdown of the technology in a video you can watch below (starting at roughly the 10 minute mark).
Watch Paris Hilton Get Ready for a Date Night With “Tom Cruise” in Deepfake Video
Fans certainly didn’t just cruise past this video on their FYPs.
Paris Hilton had plenty of TikTok users fooled when she posted a video on June 14 of her and what appeared to be Tom Cruise getting ready for a night out at his Top Gun movie premiere.
In her video, a man who strikes an uncanny resemblance to Tom walks out of the closet while putting on his blazer to tell Paris, “I don’t want to be late to this premiere.”
The This is Paris podcast host then appears in a sparkling gown, telling her date, “We should always run fashionably late—it’s your night.”
And after exchanging a few compliments, the pair embraced in front of the mirror to admire their red carpet looks.
“Do you think people are really gonna believe that we’re a couple?” Paris asked, to which “Tom” quipped, “I think most people will believe anything.”
Paris Hilton Through the Years
And most people did! The TikTok video sent the Internet into a frenzy. One user commented, “I’m so confused, but invested.”
Meanwhile, another user started to catch on, writing, “Wait is this the real Tom Cruise or the computer generated one?”
Paris eventually cleared the air by commenting that she was “Sliving it up with @deeptomcruise”—which is the TikTok account for impersonator Miles Fischer. The content creator is known to use deepfake technology to digitally place Tom Cruise’s face onto his own.
In a December 2021 interview with Today, Miles reflected on his use of the face-swapping tech.
“As I find myself the unofficial face of this deepfake movement, it’s important to learn and I’m fascinated by this,” Miles said. “This is the bleeding edge of technology.”
In real life, Paris is married to author Carter Reum. Meanwhile, Tom Cruise, who was formerly married to Katie Holmes from 2006 to 2012, Nicole Kidman from 1990 to 2001 and Mimi Rogers from 1987 to 1990, has not confirmed a new romance.
But in the TikTok world, it appears these two pop-culture icons are “together.” At the end of their shockingly realistic spoof, Paris and deepfake Tom agreed to skydive into their date night.
How did Paris feel about their adventure? She ended the video by channeling the real action star and saying, “I do my own stunts.”
While the video has racked up over 10.5 million views, it remains to be seen if the real Tom Cruise has seen this incredible stunt.
Facebook, YouTube, and Twitter remove Zelensky deepfake
The deepfake spread widely online Wednesday, as noticed earlier by Vice’s Motherboard. In the video, which CNN Business has reviewed, Zelensky appears to stand behind a presidential podium and in front of a backdrop, both of which feature the Ukranian coat of arms. Wearing a green shirt, Zelensky speaks in Ukranian, appearing to tell Ukranians to put down their weapons in the weeks-old war against Russia.
Deepfakes — which combine the terms “deep learning” and “fake” — are persuasive-looking but false video and audio files. Made using cutting-edge and relatively accessible AI technology, they aim to show a real person doing or saying something they did not. Experts have long been concerned that, as they improve, they would be used to spread misinformation.
In a series of posts on Twitter Wednesday afternoon, Meta’s head of security policy, Nathaniel Gleicher, wrote that the company spotted and removed the video earlier that day. “We’ve quickly reviewed and removed this video for violating our policy against misleading manipulated media, and notified our peers at other platforms,” he wrote.
YouTube spokesperson Ivy Choi said the video and reuploads of it have been removed from the platform because it violates the company’s misinformation policies. “We do allow this video if it provides sufficient education, documentary, scientific or artistic context,” Choi said in a statement.
A Twitter spokesperson said the company is tracking how the video is shared across the social network, and has taken “enforcement action” in cases where it violates company rules (such as its synthetic and manipulated media policy, which forbids users from sharing altered content that may confuse people or lead to harm; in some cases, Twitter may label tweets containing misleading media to give users more context).
While the video doesn’t look tremendously doctored, there are some telltale signs that the video is not what it appears to be. And Zelensky himself appeared in a video posted to an official Ukraine defense account on Twitter, saying he is continuing to defend Ukraine and refusing to lay down weapons against Russia.
Hany Farid, a professor at the University of California, Berkeley, and digital forensics expert, pointed out several of the obvious signs that the video is a deepfake. First, it’s a low-quality, low-resolution recording; this is a common trick to hide the distortions created when making a deepfake, as our brains tend to be more forgiving of glitches in low-quality videos. Second, the Zelensky in the video looks straight ahead without moving his arms throughout the clip — it’s very tricky to make a convincing deepfake that includes head motions and hands moving in front of the face. Third, there are little visual inconsistencies in the video, he pointed out, that occur during the process of making a deepfake, which is created a single frame at a time. Though Zelensky’s voice is harder for Farid to comment on, in part because he doesn’t speak Ukranian, he said it sounds a bit off to him.
The video comes weeks after the official Facebook account for Ukraine Land Forces posted a warning that such videos of Zelensky may appear. “Be aware – this is a fake!” the account wrote, soon adding, “Rest assured – Ukraine will not capitulate!” That warning was accompanied by an image that appeared to show Zelensky in a similar shirt as what appeared in the deepfake video, in front of the same backdrop and behind the same podium.
As of Wednesday afternoon, the video could still be found online, such as in some posts CNN Business spotted on Twitter and YouTube in which users made clear that it was a deepfake.
While Farid doesn’t think the video fooled people, he thinks it “muddies the information waters,” making it harder for anyone to trust what they see.
“Casting doubt on what you see and hear and read is a very powerful weapon in the information war and deepfakes are now playing a role in that,” Farid said.
how a Catherine Keener deepfake dishonors the ’90s icon.
This post contains spoilers for The Adam Project.
The villain in Netflix’s The Adam Project is meant to seem like a familiar face, but there’s something fundamentally wrong with them—their face, specifically. In the new time-travel thriller, Ryan Reynolds zaps back to the early 21st century to stop the invention of the technology that will eventually ruin the world. The problem isn’t the tech itself, which was invented by his late father shortly before his death. The problem is the unscrupulous hands it will eventually fall into—the hands of his dad’s ruthless business partner, Maya Sorian (Catherine Keener). In the future, she’s a sci-fi overlord, swathed in a black cape that drapes over her like a shroud. But when the plot rewinds to 2018, we get a look at a younger Maya, a sleek entrepreneur in shimmering blouses and an all-business ponytail. The movie’s loosey-goosey approach to temporal paradox allows the two Mayas to meet face to face, and the same de-aging technology deployed in the Marvel Cinematic Universe allows both versions of Maya to be played by Keener herself. But having grown up watching—and, let’s be honest, swooning over—Keener’s performances in the 1990s, The Adam Project’s deepfake feels unnervingly off.
It’s not just, as plenty of viewers noted online the weekend after The Adam Project’s debut, that the effects are lousy and Keener’s dialogue is out of sync with the younger Maya’s movements—or even that the CG-ed version doesn’t match the ample photographic record of what Catherine Keener actually looked like in her thirties. The blandly prettified, smooth-as-plastic Maya feels like a rebuke of Keener’s idiosyncratic screen self, a low-key desecration of one of indie film’s most important icons.
To watch American independent cinema 30 years ago was to be in love with Catherine Keener, or to want to be her. I first remember encountering Keener in 1996’s Walking and Talking, the first of five movies she made with the writer-director Nicole Holofcener. When she’s introduced, Keener’s sitting alone in a coffee shop, writing in a notebook, when her childhood friend (Anne Heche) bounces in to join her. Heche’s character has barely taken a seat when the waiter offers her some coffee, which prompts Keener to clear her throat and add that she’d like some, too—her tone suggesting the long-rehearsed slight of being overshadowed by her more self-assured blonde companion. But though she’s introduced as a wallflower, with rumpled hair and a loose-fitting t-shirt haphazardly tucked into her jeans, there’s no risk of Keener fading into the woodwork: From that first moment, she radiates keen intelligence and a ferociously active mind. As the waiter grudgingly takes her order, Keener shoots him a quick thanks-for-noticing smile, and as his back turns, she sticks out her tongue and pants like a dog. Maybe she’s mocking her own desperation or his obtuseness, or maybe she’s just firing off excess neural energy so her head doesn’t explode. But it’s a moment of pure communion with the camera, passing so quickly and unremarked-upon that you’d miss it if you were looking anywhere else. (You weren’t.)
Keener wasn’t just the thinking person’s sex symbol. She was your smartest friend, your most kind-hearted ex, the person you could trust to tell it to you straight, even if you might not like what you heard.
Especially in the movies she made with Holofcener, Keener’s characters always seem like they’re enjoying a private joke, even if the joke didn’t start out that way. She begins the cycle as the personification of Generation X: overqualified for her menial job, yet too suspicious of success to push for anything better. In Lovely & Amazing, Keener’s character bumps into an old high-school classmate and expresses shock that her childhood friend is already a practicing pediatrician. The friend, nonplussed, says, “We’re 36,” and Keener responds, “Yeah, but not 36 36.” In Holofcener’s later movies, she finds a career niche, whether it’s writing screenplays or running a vintage furniture boutique, but the professional progression doesn’t give her any firmer sense of security. She’s always adrift, restless, unsure if the problem is the world or herself. Happiness is for the simple-minded, not someone who can always anticipate the next crisis coming around the corner.
In a different era, Keener’s husky voice and deadpan demeanor might have made her a Hollywood star, the natural-born version of what Howard Hawks molded Lauren Bacall into in The Big Sleep. But Keener had no interest in going that route, often refusing to do interviews or be the subject of profiles, leaving Entertainment Weekly to pay her the backhanded compliment of praising her “unusual beauty.” In fact, she’s always been beautiful, which Holofcener acknowledged in Lovely & Amazing by making Keener’s struggling amateur artist a former homecoming queen. But her characters rarely seemed to take that to heart, too consumed with their inner failings to take stock of, let alone exploit, their outer radiance. “You’re really pretty,” Kevin Corrigan’s video store clerk tells her in Walking and Talking. “You look like you need to hear it.”
Spike Jonze’s Being John Malkovich, from 1999, is the rare occasion where Keener’s character seems in full command of her own magnetism. When John Cusack’s shambling professional puppeteer, who works in the same drab office building as Keener, makes a stammering try at asking her out on a date, she fires back a shriveling, “If you ever got me, you wouldn’t know what to do with me.” But it’s clear that this is a role that she’s playing; she steps into it the way the film’s characters briefly insert themselves into the body of the titular movie star. Keener’s costumes here—usually all black or white with occasional monochrome separates—suggest that the film wants us to see her archetypically: the unattainable angel or sexual siren. But either way, she’s irresistible. In the end, Cusack gives up his own corporeal existence just for a chance to keep looking at her.
Focus Features
In the 1990s, Keener wasn’t just the thinking person’s sex symbol, although she was called that often enough. She was your smartest friend, your most kind-hearted ex, the person you could trust to tell it to you straight, even if you might not like what you heard. She stuck to her guns even when, as it often did, it meant ending up alone, and she never fooled herself into thinking the world had gotten better just because her place in it improved. And it’s because of what she meant back then that it feels like the last decade’s done her so dirty.
Keener, who turned 40 in the year 2000, was still getting plum parts well into the new millennium; she was the woman Steve Carrell gets his shit together for in The 40-Year-Old Virgin and was nominated for an Oscar as Capote’s Harper Lee. But things started to turn with 2013’s Captain Phillips, a movie which fostered the best performance of Tom Hanks’ career but gave Keener a glorified his nondescript wife, a role so generic and functional, not even she could make it interesting. In 2017’s Get Out, Jordan Peele cast her perfectly (and cannily) as a wealthy liberal whose outward benevolence could quickly turn ice-cold. (In Please Give, she’s so riddled with privileged guilt that she offers her restaurant leftovers to a Black man on the street; it turns out he’s just waiting for his table.) But a sourness set in with 2018’s The Incredibles 2, in which she voices a wealthy tech mogul so consumed with resentment toward superheroes, she’s willing to commit mass murder to discredit them.
The Adam Project’s Maya is a similarly embittered spinster. Future Maya, played by the normal-looking 62-year-old Keener, is an imperious ruler with a legion of anonymous warriors at her command, but her domination hasn’t brought her happiness. When she zips back to the past to instruct her younger self on how to illegally seize control of the company, she doesn’t even bother to conceal her contempt for the naïve 30 year old who still thinks she can have a career and a life at the same time. “Where are you going?” Elder Maya taunts, as Younger Maya tries to exit their conversation. “Seeing someone? No, you’re not. You’re too busy. Thing is, you always will be. This company is all you will ever have. It is your personal life. It’s your family.” (Needless to say, The Adam Project doesn’t present its male characters with the same binary. Ryan Reynolds’ future traveler remembers his scientist dad as an absent father, but he’s proven wrong: The old man may have worked too hard, but he always had time for a game of catch.)
The deepfaked Maya doesn’t look like anything the young Catherine Keener, but more importantly, she doesn’t feel like her. There’s no crackle of thought behind her eyes, nothing to set her apart from the horde of identically attired tech whizzes one of her ’90s characters would have scoffed at as they sped past the coffee shop on their way to work (or worse, the gym). The embodiment of Gen X ambivalence has become a smoothly tooled success robot, every errant thought, every melancholy half-smile blasted away by a program designed to produce an illusion of life. She’s not an unusual beauty, just a usual one. And as it turns out, a digitally perfected Catherine Keener is no Catherine Keener at all.
Lucasfilm hires deepfake YouTuber who fixed Luke Skywalker in The Mandalorian
Shamook’s deepfake version of Mark Hamill as Luke Skywalker in The Mandalorian.
Shamook via YouTube
The season 2 finale of The Mandalorian saw the titular lone bounty hunter complete his mission by delivering Grogu (formerly known as Baby Yoda) to the open arms of a Jedi. But not just any Jedi. A young Luke Skywalker showed up to the excitement, then disappointment of fans who raised eyebrows at his VFX-heavy look.
Enter the YouTuber known as Shamook, whose The Mandalorian deepfake, published in December, has earned nearly 2 million views for improving the VFX used to de-age Mark Hamill. It was so good Shamook then earned a new gig with Lucasfilm and its visual effects division Industrial Light and Magic.
“As some of you may already know, I joined ILM/Lucasfilms a few months ago and haven’t had the time to work on any new YouTube content,” Shamook wrote in the comment section of a recent video. “Now I’ve settled into my job, uploads should start increasing again. They’ll still be slow, but hopefully not months apart.”
Shamook said in the comments that his job title is, “Senior Facial Capture Artist.”
Lucasfilm confirmed the new hire (via IndieWire).
“[Industrial Light and Magic is] always on the lookout for talented artists and have in fact hired the artist that goes by the online persona ‘Shamook,'” a Lucasfilm representative said in a statement.
“Over the past several years ILM has been investing in both machine learning and A.I. as a means to produce compelling visual effects work and it’s been terrific to see momentum building in this space as the technology advances.”
This is Shamook’s impressive deepfake (he fixed The Irishman too).
Deepfake videos use artificial intelligence to make it appear that a person is doing or saying something they never did, as with these Tom Cruise deepfakes. In less disturbing territory, deepfakes that improve visual effects in movies appear to be a great addition to your movie-making CV.
New movies coming out in 2021: Netflix, Marvel and more
See all photos
Lucasfilm Hires YouTuber Behind ‘The Mandalorian’ Deepfake
Finally, a story about deepfakes that’s free from the moral squishiness and consent issues that plague the people creating them. On Tuesday, a spokesperson for Lucasfilm confirmed that the company had hired a popular deepfake creator as a “senior facial capture artist” under the company’s computer graphics subsidiary, Industrial Light and Magic.
The creator, a UK-based YouTuber named Shamook, had recently garnered fame in the Star wars community after one of his deepfaked videos ended up going viral. The clip was specifically designed to improve the appearance of a young Mark Hamill during The Mandalorian season two finale and has racked up more than two million views on YouTube.
And it’s easy to see why—looking at Shamook’s deepfake work alongside the work of Industrial Light and Magic (the Lucasfilm division responsible for these sorts of CG graphics), the hobbyist’s work is arguably better than that of the pros.
Apparently, the clip was good enough to convince the Industrial Light and Magic team to recruit Shamook for the Mandalorian’s upcoming seasons. “Over the past several years, ILM has been investing in both machine learning and A.I. as a means to produce compelling visual effects work,” a spokesperson said in a statement to Indiewire about the hire. They added that the company is “always on the lookout for talented artists,” and that it’s “it’s been terrific to see momentum building in this space as the technology advances.”
Shamook hinted to the hire earlier this month in a comment left on his YouTube channel noting that since joining up with Lucasfilm earlier this year, he hasn’t “had the time to work on any new YouTube content,” on his channel, though he also added that now that he’s “settled” into the role, “uploads should start increasing again.”
While his Luke Skywalker deepfake clip might have been his big break, Shamook’s been using deepfake technology to create all sorts of content since his channel first debuted in 2018. He’s used his skills to put a deepfaked Will Smith into the Matrix franchise, to smooth over Henry Cavill’s face during his appearance in the 2017 Justice League film, and de-age a somewhat wrinkly Robert De Niro. Hopefully, in this new role, we’ll be seeing a lot more of Shamook’s work on the big screen.
G/O Media may get a commission
Lucasfilm hires YouTuber who used deepfake to improve ‘The Mandalorian’
Luke Skywalker’s CGI face in the character’s The Mandalorian cameo was met with a lot of criticism, and fans even tried to fix the scene with various tools and programs. One of those fans did so well, Lucasfilm has hired him to help it ensure its upcoming projects won’t feature underwhelming de-aging and facial visual effects. That fan is a YouTuber known as Shamook, who uses deepfake technology to improve upon bad CG effects and to put actors in shows and movies they never starred in.
In the comments section of a video that replaces Christian Bale with Robert Pattinson as the Batman in Christopher Nolan’s film, Shamook wrote that he joined Lucasfilm/Industrial Light & Magic (ILM) a few months ago. When asked what his role within the company is, he said his official title is “Senior Facial Capture Artist.” The studio has confirmed the hire with IndieWire, telling the publication that it’s always on the lookout for talented artists. A representative said in a statement:
“Over the past several years ILM has been investing in both machine learning and A.I. as a means to produce compelling visual effects work and it’s been terrific to see momentum building in this space as the technology advances.”
In addition to working on a deepfake version of Luke in The Mandalorian, he also deepfaked Tarkin’s and Leia’s appearances in Rogue One. Shamook’s videos don’t always show the most realistic results, but the great ones like Luke’s truly look impressive. Lucasfilm could his technical know-how to make sure de-aged characters and CG faces won’t take us straight to uncanny valley anymore.
All products recommended by Engadget are selected by our editorial team, independent of our parent company. Some of our stories include affiliate links. If you buy something through one of these links, we may earn an affiliate commission.
Lucasfilm hires deepfake YouTuber who fixed The Mandalorian
Shamook’s deepfake version of Mark Hamill as Luke Skywalker in The Mandalorian.
Shamook via YouTube
The season 2 finale of The Mandalorian saw the titular lone bounty hunter complete his mission by delivering Grogu (formerly known as Baby Yoda) to the open arms of a Jedi. But not just any Jedi. A young Luke Skywalker showed up to the excitement, then disappointment of fans who raised eyebrows at his VFX-heavy look.
Enter the YouTuber known as Shamook, whose The Mandalorian deepfake, published in December, has earned nearly 2 million views for improving the VFX used to de-age Mark Hamill. It was so good Shamook then earned a new gig with Lucasfilm and its visual effects division Industrial Light and Magic.
“As some of you may already know, I joined ILM/Lucasfilms a few months ago and haven’t had the time to work on any new YouTube content,” Shamook wrote in the comment section of a recent video. “Now I’ve settled into my job, uploads should start increasing again. They’ll still be slow, but hopefully not months apart.”
Shamook said in the comments that his job title is, “Senior Facial Capture Artist.”
Lucasfilm confirmed the new hire (via IndieWire).
“[Industrial Light and Magic is] always on the lookout for talented artists and have in fact hired the artist that goes by the online persona ‘Shamook,'” a Lucasfilm representative said in a statement.
“Over the past several years ILM has been investing in both machine learning and A.I. as a means to produce compelling visual effects work and it’s been terrific to see momentum building in this space as the technology advances.”
This is Shamook’s impressive deepfake (he fixed The Irishman too).
Deepfake videos use artificial intelligence to make it appear that a person is doing or saying something they never did, as with these Tom Cruise deepfakes. In less disturbing territory, deepfakes that improve visual effects in movies appear to be a great addition to your movie-making CV.
New movies coming out in 2021: Netflix, Marvel and more
See all photos
UPDATE 2-China regulators held talks with Alibaba, Tencent, 9 others on ‘deepfake’ tech
Bloomberg
Kids as Prey Spur U.S. Outrage, Bid to Cut Social Media’s Shield
(Bloomberg) — The teen was in high school when his secret spilled onto the internet, driving him to consider suicide: classmates were viewing sexual images of him and a friend on Twitter that child pornographers had duped him into sending.The videos remained visible for more than a week as the teen and his mother pleaded with Twitter Inc. to block the material, according to a lawsuit filed on the teen’s behalf. The complaint alleges the company acted only after the images drew 167,000 views and leering comments from Twitter users, with some remarking on how young the pictured victims appeared to be.Targets of online exploitation and harassment say they sometimes face indifference from platforms that operate under protection of a decades-old U.S. law that limits liability for content their users post online. The law has drawn protests from Republicans and Democrats who allege it has been used by the platforms to mishandle political speech. Now, child advocates and families say the provision has permitted companies to dodge responsibility for online harassment and even sexual exploitation. “Things like this happen all the time,” said Fordham University law professor Olivier Sylvain.The law, he said, “poses a real obstacle” for those pressing social media sites to remove material.That’s led privacy advocates, politicians and even parents of murdered children who’ve been trolled to urge Congress to restrict or do away with the legal shield, known by its chapter heading in the Communications Decency Act of 1996: Section 230.Read more about Section 230: Tech’s Liability Shield Under Fire: 26 Words and What’s at Stake Section 230: The 26 Words That Helped Make the Internet a Mess Why ‘Section 230’ Is Nub of Fights Over Online Speech: QuickTakeThe issue gained prominence during the 2020 elections when President Donald Trump and other Republicans said, with scant evidence, that it let the websites suppress conservative speech — something the sites denied. Democrats, in turn, blame the provision for an unchecked flood of misinformation about candidates or Covid-19. President Joe Biden while a candidate called for repealing Section 230. More recently his Commerce Secretary Gina Raimondo spoke of revising the law.The companies say they are doing what they can to take down offensive content, a task made difficult by the huge volume of posts. In the first half of 2020, Twitter suspended 438,809 accounts for displaying material that sexually exploited children, according to a filing in the lawsuit brought on behalf of the teen who was harassed. But, the company said, it’s not possible to remove all offending content from the hundreds of millions of tweets daily from more than 190 million users.Still, Twitter asked the court to dismiss the youth’s case, saying in the March 10 filing that “there is no legal basis for holding Twitter liable.” Under Section 230, “internet platforms are immune from suit based on the failure to remove offensive third-party content,” the company wrote.“No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.” —Section 230 of the Communications Decency Act of 1996.The issue heated up when Twitter permanently banned Trump for breaking its rules against glorifying violence after the Jan. 6 assault on the U.S. Capitol.Less prominent in the debate have been private victims, such as the youth whose photos were still circulating even after his parents say they brought it to the attention of Twitter.“It was deeply shocking, and traumatizing,” said Peter Gentala, an attorney for the National Center on Sexual Exploitation. The nonprofit group, along with two law firms and the teen’s mother, filed the lawsuit on behalf of the youth, identified as John Doe in the filings.“John Doe is in the difficult and sad position of looking for accountability because Twitter didn’t follow its own policy, or even the law,” Gentala said. “When you hear the largest companies in the world say, ‘We can’t be held responsible for this,’ it’s small wonder you see consensus building among lawmakers” to consider changes.Twitter declined to comment on specifics of the lawsuit beyond its filings, and said in an email that it has “zero-tolerance for any material that features or promotes child sexual exploitation.” Proprietary Tools“We aggressively fight online child sexual abuse and have heavily invested in technology and tools to enforce our policy,” the company said in the statement. Twitter says it uses “internal proprietary tools” to identify child sexual exploitation that have been used to close thousands of accounts.Twitter also offers an online form for reporting online child sexual material, and says that in most cases consequences for violating its ban on such material is immediate and permanent suspension. The company says it reports offensive posts to the National Center for Missing & Exploited Children, a private non-profit group that works to help find missing children and reduce child sexual exploitation.Similarly, Facebook says it uses technology to find child exploitative content and detect possible inappropriate interactions with children.“In addition to zero-tolerance policies and cutting-edge safety technology, we make it easy for people to report potential harms, and we use technology to prioritize and to swiftly respond to these reports,” Facebook said on a web page describing its policies.In Congress, lawmakers have jousted over Section 230’s effect on political speech. Now there are nearly two dozen legislative proposals to reform Section 230, according to a count by the policy group Public Knowledge. One Senate bill aims to hold social media companies accountable for enabling cyber-stalking and targeted harassment.“Section 230 has provided a ‘Get Out of Jail Free’ card to the largest platform companies even as their sites are used by scam artists, harassers and violent extremists to cause damage and injury,” said Virginia Democratic Senator Mark Warner, a sponsor of the measure announced Feb. 5 called the Safe Tech Act.The problem is widespread. Four in 10 U.S. adults say they’ve experienced online harassment, according to the Pew Research Center. Most Americans are critical of how social media companies address online harassment, the center found.Targets have included parents and families of the victims at Sandy Hook Elementary School, where 20 pupils and six teachers and staff members were slain by a gunman in 2012. Accusations quickly arose that the attack was a hoax and that Lenny Pozner, father of the youngest victim, 6-year-old Noah, had faked the event.Pozner became a target for conspiracy theorists. He said social media platforms ignored his requests to remove content, leading him to form the HONR network that began by helping to organize campaigns by volunteers to flag false and defamatory posts, bringing the content to websites’ attention for potential action. Today the site has direct relations with major platforms such as Facebook and Google, and can bring cases of people illegally harassed online to operators’ attention for possible removal of the content, Alexandrea Merrell, executive chair of the HONR board, said in an interview. Twitter hasn’t joined the initiative, Merrell said. Twitter declined to comment about its participation. Still, the HONR website laments “the apathetic and inconsistent response” by the platforms to requests to remove harmful material. The companies need better procedures to deal with misuse of the platforms, the group says.Pozner learned to use copyright claims to his family images to force sites to remove content. He’s also suing Alex Jones, host of the conspiracy website InfoWars, who had derided the shooting as fake and possibly staged by the government. The lawsuit filed in 2018 is set to come to trial this summer, Merrell said.“You can say that Sandy Hook never happened, but you can’t say that Lenny Pozner is a crisis actor who took money to pretend his son was murdered at Sandy Hook,” Merrell said. “That is defamatory.’”“Social media has been largely apathetic, simply because they can,” Merrell said.Social media sites operate in an environment shaped by years of jurisprudence.“Courts have stretched Section 230’s legal shield far beyond what its words, context, and purpose support,” Danielle Citron, a professor at the University of Virginia Law School, told lawmakers at a 2019 hearing.“It has led to the rise of social media companies like Facebook, Twitter, and Reddit. But it also has subsidized platforms that encourage online abuse,” Citron said. “It has left victims without leverage to insist that platforms take down destructive activity.”Section 230 protects platforms from lawsuits over hosting speech. At the same time, the so-called Good Samaritan part of Section 230 lets platforms weed out some speech, for instance to prevent children from being exposed to adult content, or to suppress abusive online behavior.It’s still illegal for platforms to knowingly host illegal content, such as sexually explicit conduct involving a minor.In the case of John Doe, the boy and his mother used Twitter’s online reporting form to say it was hosting illegal child pornography. The material at issue was extracted when the youth responded to an online request that he thought was from a 16-year-old girl. Once the first images were sent, the hook was set for blackmail to produce more images including those involving a friend.Twitter in automated messages (“Hello, thanks for reaching out,” began one) said it didn’t see a problem, according to the lawsuit. The material was blocked only when the family turned to a personal connection with a law enforcement official, who reached out to the company.According to the lawsuit, the youth complained to Twitter on Jan. 21 and the images weren’t removed until “on or about” Jan. 30. The platform is designed to help its users disseminate material quickly to large numbers of people, and its safeguards don’t work to quickly block illegal child pornography, according to the lawsuit.Twitter said it works to protect children.“Our dedicated teams work to stay ahead of bad-faith actors and to ensure we’re doing everything we can to remove content, facilitate investigations, and protect minors from harm — both on and offline,” the company said in an emailed statement.For more articles like this, please visit us at bloomberg.comSubscribe now to stay ahead with the most trusted business news source.©2021 Bloomberg L.P.