nov21
NÂO SÂO TÂO PERIGOSOS COMO SE PENSAVA?
Researchers from the Massachusetts Institute of Technology (MIT) have put out a new report investigating whether political video clips might be more persuasive than their textual counterparts, and found the answer is... not really. (...) To gauge how effective this tech would be at tricking anyone, the MIT team conducted two sets of studies, involving close to 7,600 participants total from around the U.S. Across both studies, these participants were split into three different groups. In some cases, the first was asked to watch a randomly selected “politically persuasive” political ad (you can see examples of what they used here), or a popular political clip on covid-19 that was sourced from YouTube. The second group was given a transcription of those randomly selected ads and clips, and the third group was given, well, nothing at all since they were acting as the control group. The result? “Overall, we find that individuals are more likely to believe an event occurred when it is presented in video versus textual form,” the study reads. In other words, the results confirmed that, yes, seeing was believing, as far as the participants were concerned. But when the researchers dug into the numbers around persuasion, the difference between the two mediums was barely noticeable, if at all. LINK + https://www.pnas.org/content/118/47/e2114388118
ATUALIZAÇÃO
APr21
Os deepfakes chegaram cedo demais??
Mar21
The US Intelligence Community this week released an unclassified version of its report on foreign interference in the 2020 Federal elections. The investigation found no evidence of foreign attempts to manipulate vote counts or other “technical aspects” of the election. That said, there were some significant foreign efforts to influence US voters. Here's how the influence horse race turned out. https://thecyberwire.com/newsletters/disinformation-briefing/3/11
ARTIGO de 9/9/20 (NAO INCLUÍ por não conhecer)
CONTEXTO FALSO Deepfake videos are proliferating on social media and the internet, and most of them are focused on politics and the coming U.S. election, according to a study by an AI-powered “synthetic media” hunting startup. “It’s far worse than you might think,” Jean-Claude Goldenstein, CEO of CREOpoint, said in a statement. “At a time of political instability paired with rapid technology changes in video manipulation and algorithmic amplification, we’re potentially barreling toward a catastrophically impacted election in just weeks.”
VERDADE? 60% of deepfake videos now target politics.
CONTEXTO “I’ve been analyzing the impact of digital deception since the aftermath of the 2016 election, and I view deepfakes as among the most horrific threats to our democracy,” said former U.S. Federal Elections Commission Chair Ann Rave https://www.forbes.com/sites/johnkoetsier/2020/09/09/fake-video-election-deepfake-videos-grew-20x-since-2019/?sh=edca3ef148ca
Mar21
PERIGO EXISTE, FBI AVISA
The FBI warned in an alert Wednesday that malicious actors “almost certainly” will be using deepfakes to advance their influence or cyber-operations in the coming weeks. The alert notes that foreign actors are already using deepfakes or synthetic media — manipulated digital content like video, audio, images and text — in their influence campaigns. “Foreign actors are currently using synthetic content in their influence campaigns, and the FBI anticipates it will be increasingly used by foreign and criminal cyber actors for spearphishing and social engineering in an evolution of cyber operational tradecraft,” states the alert obtained by CyberScoop. https://www.cyberscoop.com/fbi-foreign-actors-deepfakes-cyber-influence-operations/
FEVE21
MEMES mais faceis Why it matters: For years, there's been growing concern that deepfakes (doctored pictures and videos) would become truth's greatest threat. Instead, memes have proven to be a more effective tool in spreading misinformation because they're easier to produce and harder to moderate using artificial intelligence.
"When we talk abut deepfakes, there are already companies and technologies that can help you understand their origin," says Shane Creevy, head of editorial for Kinzen, a disinformation tracking firm. "But I'm not aware of any tech that really helps you understand the origin of memes." https://www.axios.com/memes-misinformation-coronavirus-56-2c3e88be-237e-49c1-ab9d-e5cf4d2283ff.html
amr21 Ele usou um software de código aberto gratuito chamado DeepFaceLab como base para seus deepfakes e então empregou um poderoso hardware de processamento gráfico para rodar um sistema de inteligência artificial que foi treinado com mais de 13.000 imagens de Tom Cruise de todos os ângulos imagináveis do ator. Chris então criou detalhes adicionais com um conjunto de dados menor, de 5.000 a 6.000 imagens adicionais do ator. O vídeo demorou dois meses para ser concluído. - https://www.ofuxico.com.br/noticias-sobre-famosos/deepfake-de-tom-cruise-demorou-2-meses-para-ser-feito/2021/03/09-398102.html
FEV21
FRANÇA Les «deep fake», une menace pour la campagne présidentielle de 2022. L’exécutif redoute une campagne entachée par la multiplication de «faux» quasi indétectables. https://www.lefigaro.fr/politique/les-deep-fake-une-menace-pour-la-campagne-presidentielle-de-2022-20210223TRABALHO====================================
DEez20
Mirando
atrás, estos temores fueron exagerados. (...) Aunque las predicciones más
terribles sobre los deepfakes políticos no se ha hecho realidad en 2020, debemos
analizar su evolución en el contexto de los cheapfakes y otras formas de desinformación
política. Los cheapfakes ofrecen valiosas lecciones sobre los deepfakes del futuro. La pregunta, por eso, no debería ser
"¿Cuándo surgirán los deepfakes políticos?" sino "¿Cómo
podemos mitigar las muchas formas en las que la desinformación visual está
modificando nuestra realidad política? Nina Schick es la autora de 'Deepfakes: The
Coming Infocalypse'. https://www.technologyreview.es/s/13044/los-deepfake-no-han-roto-la-democracia-en-2020-los-cheapfake-si
Dez20
“In terms of hate speech – one of the
hardest categories for machines to detect – AI systems are now
identifying 94.5 percent of it, according to Schroepfer. And, from the second quarter of
2019 to the second quarter of this year, the amount of hate speech that
Facebook's AI systems have identified and removed has increased five-fold.”
LINK
“The coronavirus, this is their new hoax.” That’s what Donald Trump says—or at least appears to say—in a new political ad airing in states across the country. The ad, created by a Democratic super PAC, strings together audio from various Trump speeches discussing the coronavirus. The spliced-together audio plays while a graph, representing the number of coronavirus cases in the U.S., grows exponentially in the background. The ad ends with audio and video of Trump declaring, “No, I don’t take responsibility at all.” According to the Trump campaign, the sequence at the beginning of the ad—the “hoax” part— is “false, misleading, and deceptive,” because the audio used in the ad was edited to “fraudulently and maliciously imply” that Trump called the coronavirus a “hoax.” The audio for that snippet was apparently taken from a single speech, given at a February rally in South Carolina; but several sentences—occurring after Trump said “coronavirus” and before he said “this is their new hoax”—were edited out. Last week, the Trump campaign filed a defamation lawsuit against a small TV station in Wisconsin that was running the ad Let’s look at Texas’s law, which makes it a Class A misdemeanor for a person to create and publish a “deep fake video” within 30 days of an election LINK + https://www.youtube.com/watch?v=bkMwvmJLnc0
“The data … reflects a horror-movie trope: ‘The killer is inside the house,’” writes author Peter W. Singer, a strategist and senior fellow at the New America think tank, in an essay for Defense One. “In 2016, Russia drove U.S. media narratives … then shaped online discussion via thousands of bots and trolls. … But 2020 election-related misinformation was mostly a domestic affair.” Social media companies and the U.S. intelligence community worked to cut down on the amount of misinformation that reached the public in the weeks before and the days after Nov. 3, with some success. The Cybersecurity and Infrastructure Security Agency, part of the Department of Homeland Security, sponsored a robust “Rumor vs. Reality” page on its website, which continued to be active after Nov. 3. In the days after the election, the agency also issued a blunt statement from its Election Infrastructure Government Coordinating Council executive committee and the Election Infrastructure Sector Coordinating Council as claims of rigged elections began to increase on social media “The November 3rd election was the most secure in American history,” read the statement, signed by the 10 council members. “While we know there are many unfounded claims and opportunities for misinformation about the process of our elections, we can assure you we have the utmost confidence in the security and integrity of our elections, and you should too.”
Why were deepfakes not as prevalent as feared? Simple edits and shameless falsehoods are easier to create — and worked just as well. “You can think of the deepfake as the bazooka and the video splicing as a slingshot,” Hany Farid, a University of California, Berkeley professor who specializes in visual disinformation, tells NPR. “And it turns out that the slingshot works.” “These falsehoods were consumed by audiences across the country, but unevenly, especially targeting swing states,” primarily Florida, Pennsylvania and Michigan.
“The implication is that those who would use deepfakes as part of an online attack have not yet mastered the technology, or at least not how to avoid any breadcrumbs that would lead back to the perpetrator,” writes Gary Grossman, senior vice president of the technology practice at Edelman and global lead of the Edelman AI Center of Excellence, at VentureBeat. “These are also the most compelling reasons … that we have not seen more serious deepfakes in the current political campaigns.”
The Defense Advanced Research Projects Agency has two programs underway to improve the detection of deepfakes: The Media Forensics, or MediFor, program is working on algorithms that can let analysts know if a photo or video is faked and how it was done; the Semantic Forensics, or SemaFor, program works to develop additional algorithms that can better identify deepfakes. “Both SemaFor and MediFor are intended to improve defenses against adversary information operations,” states a Congressional Research Service report from August that provides an overview of the programs.
https://fedtechmagazine.com/article/2020/11/election-results-remained-secure-under-barrage-disinformation-altered-video
nov20
Facebook boss 'apologises' for 'Russians fixing election' in chilling Deepfake video. Artist Stephanie Lepp is using AI to create a series of uncannily convincing videos of high-profile figures 'confessing' their mistakes Artist Stephanie Lepp has used artificial intelligence to a digital “puppet” of Zuckerberg, and made it "admit" that it’s done damage both to politics and society, saying: “People on our platform have become more prone to hating others, and hating themselves. "I was naive about Russian interference in the 2016 election, and I'm still being naive about domestic interference in 2020.” https://www.dailystar.co.uk/news/latest-news/facebook-boss-apologises-russians-fixing-22970675
Where Did the Deepfakes Go? https://hyperallergic.com/596868/where-did-the-deepfakes-go/
Last week, the New York Post published a dubious story about Hunter Biden, the son of Democratic presidential nominee and former Vice President Joe Biden. The Post story claims Hunter Biden helped arrange a meeting between Joe Biden and an executive at a Ukrainian energy company Burisma in April 2015. The series of stories contained hacked materials and personal email addresses, so Twitter (TWTR) initially prevented people from posting links to the article, sending it via direct message and retweeting it. As CNN's Chief Media Correspondent Brian Stelter noted on "Reliable Sources" Sunday, "We are not talking about fully reliable sources here," referring to the New York Post's story.The story is a "manufactured scandal," Stelter said, meant to feed "whataboutism" -- an opportunity for the right-wing media to shift the conversation from Trump's record to Biden's. https://edition.cnn.com/2020/10/18/media/new-york-post-hunter-biden-reliable/index.html + https://www.pdfa.org/hunter-bidens-email-and-the-potential-for-deepfakes-with-pdf/
Latest Visual Threats Detected as of 28 October |
Latest Visual Threats Detected as of 19 October |
Nothing happened. Nothing has continued to happen. Where did our politically charged deepfake mayhem go to? Could it still happen? Is there time? With all the increasingly surreal things happening on a daily basis, would anybody even care? What happened to my US election deepfakes? The short answer is people seem to be much more taken with pornographic possibilities than bringing down Governments. According to Sensity data, the US is the most heavily targeted nation for deepfake activity. That’s some 45.4%, versus the UK in second place with just 10.4%, South Korea with 9.1%, and India at 5.2%. The most popular targeted sector is entertainment with 63.9%, followed by fashion at 20.4%, and politics with a measly 4.5%. Depending on how fierce the US election battle is fought, strange deepfake things could still be afoot at the eleventh hour. Whether it makes any difference or not is another thing altogether, and if low-grade memes or conspiracy theories are enough to get the job done then that’s what people will continue to do. Having said that: you can keep a watchful eye on possible foreign interference in the US election via this newly released attribution tracker. Malign interference campaigns will probably continue as the main driver of GAN generated imagery. Always be skeptical, regardless of suspicions over AI involvement. The truth is most definitely out there…it just might take a little longer to reach than usual. https://blog.malwarebytes.com/cybercrime/2020/10/deepfakes-and-the-2020-united-states-election-missing-in-action/
Despite people's fears, sophisticated, deceptive videos known as "deepfakes" haven't arrived this political season. But it's not because they aren't a threat, sources tell NPR. It's because simple deceptions like selective editing or outright lies have worked just fine.But amid protests over race relations in American cities, conspiracy theories about the Coronavirus crisis, tension over President Trump's Supreme Court pick, and a contentious presidential race, few deepfakes have been used this political season. One notable exception was a faked video showing former Vice President Joe Biden sticking his tongue out, which was tweeted out by the president himself. "That actually was manipulated using deep learning-based technology," said Lindsay Gorman, the Emerging Technologies Fellow at the Alliance for Securing Democracy, about the Biden video. "And I would classify that as a deepfake." https://www.npr.org/2020/10/01/918223033/where-are-the-deepfakes-in-this-presidential-election
A Voting Rights Group Made Deepfakes of Dictators Warning About the Death of Democracy . The campaign focuses on voter suppression and misinformation that threatens open elections https://www.adweek.com/programmatic/dictators-deepfakes-death-of-democracy-representus/
The Trump campaign and its surrogates have seized on Democratic nominee Joe Biden's age and have been painting him as mentally unfit for the presidency. Videos of Biden falling asleep during an interview, misspeaking about the dangers of "Joe Biden's America" and appearing lost during a campaign event have bolstered the belief, particularly among Trump supporters, that Biden is in cognitive decline.
There's just one problem: None of these videos are what they seem, and some of the events depicted didn't happened at all. Technological developments have made it easier for people to produce seemingly real videos that are anything but. These deceptively altered videos have become a major element of disinformation campaigns that wield falsehoods in an effort to sway voters. https://kiowacountypress.net/content/faked-videos-shore-false-beliefs-about-bidens-mental-health
And then, at 8:25:50 pm ET, the president retweeted an account he had never retweeted before. The account had posted a video of former Vice President Joe Biden, crudely and obviously manipulated to show him twitching his eyebrows and lolling his tongue. The caption read: “Sloppy Joe is trending. I wonder if it’s because of this. You can tell it’s a deep fake because Jill Biden isn’t covering for him.” https://www.theatlantic.com/ideas/archive/2020/04/trumps-first-deepfake/610750/ + https://www.dailymail.co.uk/news/article-8260455/Trump-shares-deep-fake-GIF-Joe-Biden-sticking-tongue-series-late-night-posts.html The president’s use of this gif is already coming in for criticism. Writing in the Atlantic, David Frum noted the significance of the president’s retweet: “Instead of sharing deceptively edited video—as Trump and his allies have often done before—yesterday Trump for the first time shared a video that had been outrightly fabricated.” https://www.lawfareblog.com/deepfake-iphone-apps-are-here
No comments:
Post a Comment