Regulators are using an AI system to scan websites and messaging apps to find pornography. Creators face fines and potential prison sentences.

Paulita Paupel, who runs the European branch of the adult industry trade body the Free Speech Coalition, says the crackdowns are having worrying impacts on people and their ability to share content online. “People are fleeing the country,” Paupel says. “Most major content creators have already changed their residence to other European countries, mostly Austria, Switzerland, and Cyprus.” Others have changed their marketing strategies to avoid Twitter (impacting how much money they can make), and people new to the industry may be discouraged from starting a career, Paupel says. “This is mostly affecting LGBTQI+ and BIPOC creators.”

The internet is, of course, awash with porn—from Reddit, Snapchat, and Twitter to OnlyFans, PornHub, and xVideos—with millions of people around the world involved in the industry. Globally, it is big business, generating billions of dollars every year. While there are crackdowns on pornography all around the world, Germany appears to have a particularly strong brand of enforcement in the Western world, despite being one of the highest consumers of pornography.

“Germany has been the most aggressive about suppressing speech,” says Mike Stabile, a spokesperson for the US-based Free Speech Coalition. “I think that Germany has been the most aggressive in this in its pursuit, both in terms of the scope of its laws, and then also the enforcement.” 

AI Surveillance

Since 2019, Germany’s media regulators have been developing and then using an AI system to detect online content that may run foul of the country’s laws. The artificial intelligence system, called KIVI, was developed by the North Rhine-Westphalia media authority, along with a Berlin-based private company, and is now being used by all the media authorities around Germany. 

KIVI is touted as being able to scan public posts on seven social media and messaging apps—including Twitter, YouTube, TikTok, Telegram, and VK (Russia’s version of Facebook)—as well as websites on the open internet. Meta’s Facebook and Instagram, which forbid nudity, are currently not being scanned. According to North Rhine-Westphalia’s description of the tool, it can check 10,000 pages per day. Shortly after the authority started using KIVI, it said the authority’s detections “skyrocketed.”

The spokesperson for the North Rhine-Westphalia media authority says that since 2021 the authority has detected almost 5,000 “violations.” The system searches for problematic content by looking for predetermined German keywords and links, and the authority says it uses a combination of image recognition and text recognition to detect “positive” results.

Ella Jakubowska, a senior policy adviser at the civil rights nonprofit European Digital Rights (EDRi), says people’s human rights are put at risk when Big Tech companies or governments handle content moderation. “But the idea of state entities controlling what we do and do not see online seems in itself very concerning,” Jakubowska says. 

KIVI looks for multiple types of content, including political extremism and Holocaust denial, violence, and pornography. However, porn “violations” top the list, with 1,944 incidents being flagged in the past two years, according to figures shared by the North Rhine-Westphalia media authority. The spokesperson says the system flags potential violations of laws and then human investigators examine the results and decide whether any action should be taken. “KIVI protects employees from being suddenly and unexpectedly exposed to stressful content,” Plass from the Berlin authority says.

They Posted Porn on Twitter. German Authorities Called the Cops

The GOP-fueled far right differs from similar movements around the globe, thanks to the country’s politics, electoral system, and changing demographics.

The US Republican Party has become increasingly authoritarian and extreme in recent years, and it doesn’t seem likely to moderate that in the foreseeable future. Despite performing poorly in the 2022 midterms after running many candidates the public saw as too extreme, the GOP has decided to elevate and empower far-right lawmakers like representatives Marjorie Taylor Greene and Matt Gaetz. 

In Florida, books have been removed from school shelves as governor Ron DeSantis tries to reshape the public education system in his own image. Republican lawmakers around the US have passed abortion bans that put pregnant women’s lives in danger. The rights of transgender people are under attack throughout the country. 

Nearly half of Republicans say they would prefer “strong, unelected leaders” over “weak elected ones,” according to a September Axios-Ipsos poll, and around 55 percent of Republicans say defending the “traditional” way of life by force may soon become necessary. About 61 percent of Republicans don’t believe the results of the 2020 presidential election. 

Finding examples of extremism, a lust for authoritarian leaders, and general antidemocratic beliefs in America is not difficult these days—just spend a few minutes online. The question is how far down the rabbit hole the United States has gone and where it may end up in the not-too-distant future. 

“To call a party democratic—committed to democracy—they’ve got to do three basic things: They have to unambiguously accept election results, they have to unambiguously renounce violence, and they have to consistently and unambiguously break with extremists or antidemocratic forces,” says Steve Levitsky, a professor of government at Harvard University. “I think the Republican Party now fails these three basic tests.”

Levitsky says far too many Republican leaders have flirted with using violence to achieve their political goals and spread lies about the most recent presidential election. He says politicians like DeSantis appear to be experimenting with an authoritarian way of governing in their own states that could be applied at the national level should they successfully run for president. 

It’s difficult to find an apt comparison between the Republican Party and authoritarian movements that have risen elsewhere for a variety of reasons. One, Levitsky says, is that Donald Trump took over a party that has existed for nearly 170 years and made it more authoritarian. Historically, authoritarians tend to start their own parties. Another is that a relatively small percentage of the populace was able to wield such great power under Trump.

“There’s a minority of the population that’s pretty reactionary and, by a bunch of measures, fairly authoritarian in really all Western democracies,” Levitsky says. “The question is, how are they channeled into politics? What’s exceptional about the United States is that 25 percent or so was actually able to wield national power. Is MAGA comparable to far-right parties in Europe? Yeah. With the exception of maybe Golden Dawn in Greece, though, probably more openly authoritarian.”

Authoritarian movements of the past share characteristics with what we’re seeing in the US today—from Turkey and Hungary more recently to the rise of fascism in the 1920s—but the US governmental system and political parties present particular hurdles and windows of opportunity. 

Assuming democracy remains intact in the years to come, Levitsky thinks the GOP will have to eventually moderate its stance in response to changing demographics. The current extremism will not be sustainable if the party hopes to win enough elections to wield power in the future. However, Levitsky thinks any adjustments could take longer than one would hope.

“The problem is our incentives—the Electoral College, the Supreme Court, the fact that sparsely populated territories are dramatically overrepresented in our electoral system—allows the Republicans to wield a lot of power without winning national majorities,” Levitsky says. “If the Republican Party actually had to win over 50 percent of the national vote to control the Senate, to control the presidency, to control the Supreme Court, you would not see them behaving the way they’re behaving. They would never win.”

It remains to be seen whether Trump will be the Republican nominee in the 2024 presidential election, but there’s clear evidence that the effects of his actions wouldn’t simply disappear if he wasn’t controlling the party. A lot of Americans have been radicalized since he first took office, and it’s not easy to roll that back.

“I think the consensus is that democracy is not in the clear, and that’s because the rhetoric and actions of the GOP have emboldened their supporters to sort of accept certain behaviors that we wouldn’t have thought were in line with democracy,” says Erica Frantz, an associate professor of political science at Michigan State University. “Suddenly it’s OK to question if our elections are free and fair. Suddenly it’s OK to be provocative and suggest you might use violence if the election doesn’t go your way.”

Frantz says large sectors of the US population accept the authoritarian messaging Trump spearheaded, and that is likely going to have lasting effects. She says the fact that Trump was successfully removed from office despite his attempts to overturn the election in 2020 is a big deal, but there’s still a lot of work to be done to protect American democracy.

“I don’t think we’re going to backslide to dictatorship. The probability is higher than before Trump, but it’s still low compared to many other countries,” Frantz says. “It is very possible that we’ll muddle along for quite some time in this situation where undemocratic norms are being spouted and perpetuated by one of our main parties.”

In terms of what supporters of democracy can do in the face of an authoritarian movement, there’s no silver bullet—but there are ways to push back. Levitsky says it’s important to form large coalitions to “isolate and defeat” authoritarians, which means uniting democracy supporters on the left and the right. 

A. James McAdams, a professor of international affairs at the University of Notre Dame, says those who oppose authoritarianism need a strong message that will appeal to people who might be pulled in by authoritarian leaders. 

“If you look back historically, one of the big problems in democracies has always been that the forces of reason can’t figure out what they stand for,” McAdams says. “We’re at a point in history today in the United States and Europe where moderate parties aren’t sure what to say.”

You also need to support and strengthen democratic institutions like the courts, McAdams says. He says this is particularly important because weak courts are often part of the reason authoritarians are able to take power and chip away at democracy, such as in Latin America in the 1970s. 

“If you do have stable democratic institutions—particularly viable courts—then there’s a lot of bullshit that you can overcome,” McAdams says. “Perhaps the greatest victory for American institutions in the Trump age was that the courts weren’t overpowered.”

The Uniquely American Future of US Authoritarianism

Any mention of crypto was deliberately veiled at this year’s festival. And that strategy might catch on.

In 2022, I attended my first SXSW, and I couldn’t escape the suffocating atmosphere of crypto evangelism. A convention that touts itself as the nexus of art and technology, it seemed fertile ground for the seemingly growing NFT community. This year? I could barely find a mention of crypto. And the few who did bring it up seemed embarrassed to do so.

That strategy of non-acknowledgment might just be the future of crypto.

For the uninitiated, SXSW is an event that takes over nearly the whole city of Austin. There’s the main convention center, panels at different hotels, and concerts at the biggest stadiums. But almost every bar, club, and venue also has some tie-in party or concert. And those that don’t host official SXSW events hold unofficial ones to at least catch some of the hype.

All of that is to say, if you want to set up shop during SXSW, you can probably find a place to do so. This is what made it so bizarre that I was able to go most of the week with scarcely a hint of words like crypto, blockchain, or NFT.

In 2022, crypto blew the doors off the entire city. An outdoor venue with giant domes housed bombastic raves celebrating some little-known bunny NFTs called Flufs. (Incidentally, I attended the event last year, and the 3D images of crudely rendered rotting rabbits still occasionally haunt me.)

Blockchain Creative Labs—a division of Fox Entertainment—was a key sponsor of SXSW 2022. In 2023, BCL was nowhere to be found. Last year, there were dozens of panels advocating crypto’s benefits. This year, the word “crypto” only appears a dozen times in the SXSW event schedule. (Both “AI” and “metaverse” clocked at least 30 mentions.) Four of those were showings of a documentary about (among other things) “WallStreetBets and crypto fanatics,” and one was a panel with prominent crypto skeptic Molly White.

The minimal crypto presence at SXSW isn’t terribly surprising. Despite claims to the contrary, NFTs don’t actually work the way a lot of their advocates say they do. Many artists have rejected NFTs entirely and find them to be an external headache, rather than a useful business tool.

It’s not the first time tech has promised to revolutionize an industry only to fundamentally misunderstand the field it’s entering. (Just ask Stadia how easy it is to build a gaming platform.) But what’s notable is that crypto didn’t abandon SXSW entirely. 

Instead, it simply came in disguise.

On the expo floor, I saw a few companies that were still proud to admit they used crypto tech to insert a financial layer into an otherwise existing product. I saw a blockchain-based camera and a crypto streaming platform—both with names I’d never heard of—in tiny booths. The largest booth that prominently announced its crypto affiliation was Polkadot, a startup that “unites and secures a growing ecosystem of specialized blockchains called parachains.”

For every one of these companies, however, I saw just as many (if not more) that employed crypto while trying very hard to avoid mentioning that fact. Two companies claimed to be building the future of social media. Dig deep enough into their websites and they both offered users crypto-based incentives, but neither chose to feature crypto or blockchain tech as a selling point.

One of them, Ascend, had flyers all over the city. These made lofty—arguably impossible—promises, such as “no misinformation” (who defines what counts as misinformation?). Some of these promises were fundamentally contradictory, such as “no toxicity” and “no hate speech” but also “no centralized censorship.” It’s unclear how the company expects to reconcile many of these competing priorities, but according to its site the solution somehow involves earning “Ascend credits,” which are only described as crypto in a chart.

Another company, Arkive, launched a DAO in 2022 aimed at creating a community of members that would use NFTs and the blockchain to form a museum curated by the internet, rather than a central organization. The group met at SXSW 2023 and even put on a panel about the decentralization of art, but they downplayed the DAO and the crypto angle. Even Arkive’s own coverage of its SXSW 2023 presence barely mentions crypto.

In some cases, it’s unclear whether companies have abandoned their crypto plans or would simply prefer not to highlight them. Even expo booths for companies that are widely known for their work in crypto seemed hesitant to use any of the keywords closely associated with it. A display for _The Sandbox_—sharing a small booth with some other developers in the space—proudly touted the “metaverse” game and occasionally mentioned being a “Web3” platform. But the fact that much of the game is built around NFTs on the blockchain was somewhat obscured. 

It’s a dynamic I started to internally refer to as crypto-obfuscation. It’s not that any of these companies would refuse to acknowledge crypto, per se. When asked, many were all too happy to discuss their vision of a blockchain-based future. But they seemed to operate as though calling attention to it unprovoked was, at best, a little uncouth. At worst? An active deterrent.

Crypto has often been compared to the early internet, where the tech is exciting but not ready for normies yet. Still, no matter how cringey the internet was in its youth, there was never a time when companies avoided saying they were building a product on “the web” or “online.”

I’ll openly admit that I was deeply skeptical of crypto, even in 2022. There was already enough evidence of scams, rugpulls, disinformation, and fraud to make anyone wary of the blockchain for the next decade at least. But I felt compelled to keep my opinions a little quiet. At one crypto-themed party that year, a friend shouted quite loudly, “_NFTs SUCK!!_” And while I aspire to her energy, I also lightly shushed her for fear someone would take offense. 

This year, I felt like my skepticism had become the norm, or at least mainstream enough to express openly. Out of nearly everyone I spoke to, the few with any opinions about crypto seemed eager to share their doubts. Most simply hadn’t thought about the technology. And besides, generative AI was much more interesting to discuss.

I doubt any of this means crypto is dead or dying. The tech has been around in some form or another for over a decade, and public interest in it comes in waves. However, its subdued presence at SXSW suggests its advocates had learned a powerful lesson from the previous year: The best way to evangelize crypto outside the tech bubble is to hope you can convince people to pay no attention to the blockchain behind the curtain.

Crypto Was Afraid to Show Its Face at SXSW 2023

Plus: The “Clop” gang's ransomware spree, the DC Health Link breach comes into focus, and more.

A US House of Representatives hearing this week about the social media app TikTok did little to clarify lawmaker's specific concerns about the potential national security risks associated with the wildly popular app, but it did vividly underscore the country’s lack of federal data privacy legislation. WIRED also discovered that TikTok paid for influencers popular on its platform to attend a DC rally in support of the service ahead of the hearing. 

Meanwhile, as a possible indictment of former US president Donald Trump looms in New York state, internet users began generating AI images of Trump being arrested, but there are ways to tell that they're fake. WIRED examined the increasingly aggressive and desperate tactics of Iran's government-backed hackers amid mass protest and unrest in the country. Citizen sleuths around the world are using open source intelligence to separate fact from fiction in the mystery of who sabotaged the Nord Stream pipeline. And vulnerabilities keep showing up in ultra-popular photo cropping tools, exposing a host of cropped images all over the world where some or all of the original image can be recovered.

Plus, if you want to know what it's like to be investigated by the US Secret Service—and how to avoid that particular pleasure—we have a full account.

And there's more. Each week, we round up the security news we didn’t cover in-depth ourselves. Click the headlines to read the full stories, and stay safe out there.

People living in the Indian state of Punjab grappled with an internet shutdown for days after police imposed a connectivity blackout while searching for the Sikh activist Amritpal Singh. Singh is a member of the Sikh Waris Punjab De movement and recently evaded arrest. More than 100 of his supporters have been arrested. Punjab's 27 million inhabitants faced mobile data and SMS blocking as well as traffic filtering on certain websites and services. For example, the government appeared to have blocked access to prominent Sikh Twitter accounts, including that of poet Rupi Kaur and the nonprofit United Sikhs. “Punjab Police India continued its crackdown on Waris Punjab De elements wanted on criminal charges,” the government of Punjab said in a Facebook post on Monday. “Amritpal Singh remains a fugitive, and efforts are being made to arrest him.” Protests have erupted in Punjab and around the world over law enforcement treatment of Sikh Waris Punjab De and the internet shutdown.

A vulnerability in file transfer software from Fortra known as GoAnywhere has been repeatedly exploited by the notorious, Russia-based Clop ransomware group to target dozens or possibly more than a hundred victims in recent days. The cybercrminal group has added entries on numerous organizations to its dark web site, where Clop attempts to extort money from victims by publishing samples of data they've stolen and threatening to leak more if targets don't pay. TechCrunch confirmed on Thursday that the City of Toronto is one of the victims of the spree. “Today, the City of Toronto has confirmed that unauthorized access to City data did occur through a third-party vendor. The access is limited to files that were unable to be processed through the third-party secure file transfer system," officials said in a statement. TechCrunch has also uncovered details about problems with Fortra's response to the discovery of the vulnerability.

The company that runs the Washington DC health insurance marketplace DC Health Link suffered a breach earlier this month that exposed sensitive and personal data from tens of thousands of area customers, including from some US lawmakers and congressional staff. The information included names, email addresses, dates of birth, mail addresses, Social Security numbers, and policy details. The DC Health Benefit Exchange Authority acknowledged the breach on March 7. The entity that has claimed credit for the breach, who goes by the handle “Denfur,” posted samples of data from the attack on BreachForums. Denfur subsequently posted “Glory to Russia!” and that the “intended target was US politicians and members of US government." In an interview with CyberScoop on an encrypted chat service, Denfur claimed not to be concerned about suffering repercussions from law enforcement. “If anything, I’m more worried about my country trying to do a favour for the US and myself or group becoming a sort of bargaining chip,” Denfur said. “The current time brings uncertainty.”

The alleged “pompompurin” administrator of the popular cybercriminal public square BreachForums—the same site Denfur used against DC Health Link—was arrested in New York state earlier this month, but a new leader known as “Baphomet” had come forward, claiming to have a plan to keep the platform going. On Tuesday, though, Baphomet changed course, claiming that someone had gained access to the BreachForums backend and that law enforcement may now control pompompurin's privileged administrator accounts. “This will be my final update on Breached, as I've decided to shut it down,” Baphomet wrote. “I'm aware this news will not please anyone, but it's the only safe decision now that I've confirmed that the glowies likely have access to Poms machine.”

India Shut Down Mobile Internet in Punjab Amid Manhunt for Amritpal Singh

The interrogation of CEO Shou Zi Chew highlighted US lawmakers’ own failure to pass privacy legislation.

In one sense, today’s US congressional hearing on TikTok was a big success: It revealed, over five hours, how desperately the United States needs national data-privacy protections—and how lawmakers believe, somehow, that taking swipes at China is a suitable alternative. 

For some, the job on Thursday was casting the hearing's only witness, TikTok CEO Shou Zi Chew, as a stand-in for the Chinese government—in some cases, for communism itself—and then belting him like a side of beef. More than a few of the questions lawmakers put to Chew were vague, speculative, and immaterial to the allegations against his company. But the members of Congress asking those questions feigned little interest in Chew’s responses anyway. 

Attempts by Chew, a 40-year-old former Goldman Sachs banker, to elaborate on TikTok’s business practices were frequently interrupted, and his requests to remark on matters supposedly of considerable interest to members of Congress were blocked and occasionally ignored. These opportunities to get the CEO on record, while under oath, were repeatedly blown in the name of expediency and for mostly theatrical reasons. Chew, in contrast, was the portrait of patience, even when he was being talked over. Even when some lawmakers began asking and, without pause, answering their own questions.

The hearing might’ve been a flop, had lawmakers planned to dig up new dirt on TikTok, which is owned by China-based ByteDance, or even hash out what the company could do next to allay their concerns. But that wasn't the aim. The House Energy and Commerce Committee was gathered, it said, to investigate “how Congress can safeguard American data privacy and protect children from online harms.” And on that, the hearing revealed plenty.

For one thing, it’s clear that the attempts to isolate TikTok from its competitors—to treat it differently than dozens of other companies with atrocious records of endangering kids and abusing private data—is a pointless exercise. Asking about TikTok's propensity for surveilling its own users, Chicago congresswoman Jan Schakowsky warned Chew against using legal, typical industry practices as a defense against these wrongs. “You might say, ‘not more than other companies,’” she said, adding that she preferred not to “go by that standard.” 

OK. But why not? 

The truth is that if TikTok were to vanish tomorrow, its users would simply flock to any number of other apps that have no qualms about surveilling the most private moments of their lives and amassing, manipulating, and selling off sensitive information about them. Excluding the most serious but largely unsubstantiated allegations leveled at TikTok—that it is acting or will act in coordination with Chinese intelligence services—there wasn’t a concern about privacy raised by lawmakers Thursday that couldn't be addressed by existing legislation supporting a national privacy law. 

Ensuring that companies and the data brokers they enrich face swift reprisals for blatantly abusing user trust would have the benefit of addressing not only the accusations levied against TikTok, but deceitful practices common across the entire social media industry. 

The irony of US lawmakers pursuing a solution to a problem that’s already been solved by draft legislation—but not actually fixed due to its own inaction—wasn’t entirely lost on the members. While primarily focused on a single company, the hearing, Florida congresswoman Kathy Castor said, should really serve as a broader call to action. “From surveillance, tracking, personal data gathering, and addictive algorithmic operations that serve up harmful content and have a corrosive effect on our kids’ mental and physical well-being,” she said, Americans deserve protection, no matter the source.

The TikTok Hearing Revealed That Congress Is the Problem

The embattled social media company brought out the checkbook to ensure at least 30 of its biggest assets—creators—were in DC to help fend off critics.

Ahead of TikTok CEO Shou Zi Chew’s much-anticipated testimony in the United States House of Representatives today, the embattled tech firm conducted a full-court press on Capitol Hill. This included paying to bring TikTok influencers face-to-face with their home state lawmakers, staffers, and journalists, as well as sharing their journey with their collective audience of some 60 million followers. 

TikTok covered travel, hotels, meals, and shuttle rides to and from the Capitol for dozens of influencers, according to the creators and the company itself. Each social media star was also invited to bring a plus one—whether they flew in from Oklahoma, hopped the Acela from New York, or drove in from their suburban Washington home. TikTok spokesperson Jamal Brown confirms that “TikTok covered travel expenses for all creators and a guest.”

“Any barriers to getting here they helped cover,” says Tiffany Yu, a Los Angeles-based influencer and disability advocate tapped to speak yesterday at a highly orchestrated press conference under the Capitol’s majestic dome.

While some influencers report paying their own airfare to Washington, everyone we talked to took the free hotel. It’s unclear precisely what folks were offered as part of the trip to Washington, but seemingly everyone got one perk or another. Beyond the more than 30 influencers in attendance, along with their travel buddies, WIRED counted 10 other people who were, in one way or another, at the Capitol on behalf of TikTok.

"More than 150 million Americans, including 5 million US businesses, rely on TikTok to innovate, find community, and support their livelihoods,” spokesperson Brown says. “A US ban on TikTok could have a direct impact on the livelihoods of millions of Americans. Lawmakers in Washington debating TikTok should hear firsthand from people whose lives would be directly affected by their decisions.” 

The nearly dozen influencers WIRED spoke with did not hide the fact that TikTok brought them to Washington to support the company. (The plan was first reported by Politico and The Information.) “They took us here, but we’re not being paid,” says Jorge Alverez, a mental health advocate from New Jersey. TikTok “paid for transportation—that’s also public information.” 

Alexandra Doten, an expert in space communications who goes by @astro\_alexandra on the app, is based near Washington, DC. But she says she also received support from the company. “I got the hotel too!” she says. “I don’t know. They just shuttle me there.”

While Doten was able to meet with her congressman, Maryland’s Glenn Ivey, this week, the highlight for her was meeting astronaut-turned-senator Mark Kelly of Arizona. She also got a true taste of life at the Capitol when the state’s other senator, Kirsten Synema, sent a staffer to meet with the influencers on her behalf. It’s unclear if their goodwill tour will sway any of TikTok’s countless congressional critics, who claim the app poses a threat to US national security.

Chew is the latest in a line of Big Tech executives who’ve had to face hostile (if often uninformed) lawmakers. But unlike Meta’s Mark Zuckerberg or Alphabet’s Sundar Pichai, Chew is the only CEO who has had technology banned on US government devices or been accused of being a puppet for the Chinese Communist Party. 

TikTok’s response to lawmakers asserting that the CCP could use the app for everything from espionage to election-influence operations is to launch a PR offensive on Washington. In addition to the more than $10 million the company has dropped on lobbyists over the past two years, this week it also tapped the Los Angeles-based Hotline Agency, “a Black and queer-owned boutique creative communications agency,” to help retell its story. Hence, the company spending lavishly this week to ensure its best assets—its diverse group of creators—get private face time with Washington power brokers before members of the House Energy and Commerce Committee accuse Chew of being an operative for Chinese communists.

At TikTok’s press conference yesterday, the enthusiasm emanating from the crowd of more than 140 people was tangible; influencer stories were deeply personal. After a brutal month for the beleaguered tech firm, all the cheering and sign-waving for Yu and other speakers, including three sympathetic Democratic lawmakers, made it seem like TikTok had finally turned the tide and gotten organic momentum on its side. That tone soon clashed with the TikTok signs held by members of the crowd that had a glossy (but not too glossy) sheen indicative of a high-priced PR firm.

Much as the app’s users are addicted to its unparalleled For You feature, TikTok’s team carefully manicured this press conference to appeal to journalists’ most basic instincts. A taco influencer, with a cake served up for dessert; a nurse; a “patriot”; artists; musicians; and a model. The press corps, which made up roughly half the crowd, couldn’t look away. 

The tech company isn’t just throwing money at its image problems. It also announced new safety features ahead of today’s hearing, including acquiescing to pressure to drastically limit the time children are allowed to spend on the app daily, from unlimited to 60 minutes. 

TikTok also enjoyed the vocal support of some lawmakers. Advocates for the app say the company poses no more of a threat to America’s children than, say, Facebook did when it allegedly buried research showing Instagram was a health risk to children, according to one whistleblower.

“My question is, why the hysteria and the panic and the targeting of TikTok?” Congressman Jamaal Bowman asked yesterday’s crowd. “As we know, Republicans, in particular, have been sounding the alarm creating a red scare around China.”

But it’s not just Republicans calling for a TikTok crackdown. Senate Majority Leader Chuck Schumer has spoken out against the app, and House Minority Leader Hakeem Jeffries appointed Raja Krishnamoorthi of Illinois—one of TikTok’s loudest foes in the Capitol—as the top Democrat on Republicans’ new, high-profile select committee on China. Senator Mark Warner, a Virginia Democrat, recently introduced bipartisan legislation that could lead to a full TikTok ban in the US. 

While the anti-TikTok furor marks one of the most bipartisan policy areas in today’s divided capital, Congressman Bowman says the rush to punish TikTok is distracting lawmakers from the serious privacy concerns he has with Silicon Valley’s finest. “It poses about the same threat that companies like Facebook and Instagram and YouTube and Twitter pose, so let’s not marginalize and target TikTok,” Bowman told the crowd. “Let’s have a comprehensive conversation about legislation that we need—federal legislation—to make sure people who use social media platforms are safe and their information is secure and their information is not being shared or sold to third parties.” 

Bowman also correctly pointed out that “there are still data brokers who sell our data to other countries and businesses. In other countries, they sell to the highest bidder.” He concluded: “Let’s not be racist toward China and express our xenophobia when it comes to TikTok because American companies have done tremendous harm to American people.”

As for the influencers, they say it’s hard even to have a dialog with many members of Congress because fear of China has blinded them to what the influencers see as the beauty of TikTok’s social platform. In her remarks, influencer and Ohio native Nichole Baedri (aka @Backinthekitchenwithbae) invited the nation’s political class to join the wild ride that is TiKTok. “Connect with us on another level,” she said. Baedri further implored lawmakers to recognize the value of the app, like how it has helped marginalized individuals connect and share their stories in an impactful way. “Use the app as intended.”

When House Energy and Commerce Committee member Marc Veasey, a Texas Democrat, learned that TikTok had footed the bill for more than 30 influencers and hired a PR firm to polish its image, he says he was “not surprised”—even if his reaction suggested otherwise.

“I did not know that,” he said. “Wow. Wow.” 

During the hearing, Veasey pressed Chew on all the voter disinformation on the platform, but he says the CEO’s response was guarded and off-putting. As you might expect, Veasey’s takeaway is the opposite of what TikTok influencers and staffers were selling to lawmakers on Capitol Hill all week.

“He came here today to be very careful, instead of transparent,” Veasey says. “He came to be careful and evasive. He did not come to be forthcoming.”

TikTok Paid for Influencers to Attend the Pro-TikTok Rally in DC

Image-editing tools from Google and Microsoft contain the “aCropalypse” bug, which can reveal information users intentionally removed.

At the beginning of March, Google released an update for its flagship Pixel smartphones to patch a vulnerability in the devices’ default photo-editing tool, Markup. Since its 2018 introduction in Android 9, Markup’s photo-cropping tool had been quietly leaving data in a cropped image file that could be used to reconstruct some or all of the original image beyond the confines of the crop. Though now fixed, the vulnerability is significant because Pixel users have for years been making, and in many cases presumably sharing, cropped images that may still contain the private or sensitive data the user was attempting to eliminate. But it gets worse.

The bug, dubbed “aCropalypse,” was discovered and originally submitted to Google by security researcher and college student Simon Aarons, who collaborated on the work with fellow reverse engineer David Buchanan. The pair were stunned to discover this week that a very similar version of the vulnerability is also present in other photo-cropping utilities from a totally separate yet equally ubiquitous codebase: Windows. The Windows 11 Snipping Tool and Windows 10 Snip & Sketch tool are vulnerable in cases where a user takes a screenshot, saves it, crops the screenshot, and then saves the file again. Photos cropped with Markup, meanwhile, retained too much data even when the user applied the crop before first saving the photo. 

Microsoft told WIRED on Wednesday that it is “aware of these reports” and that it is “investigating,” adding, “we will take action as needed.”

“It was pretty mind-blowing really, it was as if lightning had just struck twice,” says Buchanan. “The original Android vulnerability was already surprising enough that it hadn’t been discovered already. It was quite surreal.”

Now that the vulnerabilities are out in the open, researchers have started uncovering old discussions on programming forums where developers noticed the odd behavior of the cropping tools. But Aarons seems to have been the first to recognize the potential security and privacy implications—or at least the first to bring the findings to Google and Microsoft.

“I actually noticed it at about 4 in the morning by total accident when I spotted that a small screenshot I sent of white text on a black background was a 5 MB file, and that didn’t seem right to me,” Aarons says.

Images impacted by aCropalypse often can’t be completely recovered, but they can be substantially reconstructed. Aarons provided examples, including one in which he was able to recover his credit card number after he attempted to crop it out of a photo. In short, there is a population of photos out there that contain more information than they should—specifically, information that someone intentionally tried to remove.

Microsoft hasn’t issued any fixes yet, but even those released by Google don’t mitigate the situation for existing image files cropped in the years when the tool was still vulnerable. Google points out, though, that image files shared on some social media and communication services may automatically strip out the errant data.

“As part of their existing compression process, apps and websites that recompress images, like Twitter, Instagram, or Facebook, delete extra data automatically from images uploaded. Images posted to sites like these are not at risk,” Google spokesperson Ed Fernandez says in a statement.

The researchers point out, though, that this is not true of all platforms, including Discord.

As a Discord user, Buchanan say he kept seeing people posting cropped screenshots, and it was really hard to not say anything before the vulnerability was publicly disclosed.

Steven Murdoch, a professor of security engineering at University College London, notes that in 2004 he discovered a vulnerability in which an older version of an image was stored in the thumbnail data for the image even after it had been altered.  

“This isn’t the first time I’ve seen this sort of vulnerability,” Murdoch says. “And I think the reason is because when software is written, it’s tested to make sure that the thing you expect is there. You save an image, you can open the image, and then you’re done. What is not checked is whether there is accidentally extra data stored.”

The thumbnail vulnerability Murdoch found in 2004 was conceptually similar to aCropalypse from a data privacy standpoint but had very different technical underpinnings because of issues in application programming interface design. And Murdoch emphasizes that while he sees aCropalypse as a problem for users whose affected photos are already out in the world, its biggest impact may come from the discussions it has raised about how to promote better security practices in API development and implementation.

“This has triggered some interesting conversations about API design and what do you do to teach people to avoid this sort of vulnerability in the future? This is not something that we train people to deal with,” Murdoch says. “It’s not one of these ‘sky is falling’ vulnerabilities, but it’s not good.”

Bug in Google Markup, Windows Photo-Cropping Tools Exposes Removed Image Data

On Thursday, Shou Zi Chew will meet a rare united front in the US Congress against the Chinese-owned social media app that has lawmakers in a tizzy.

March has been brutal for TikTok. Last week, the UK joined the US, Canada, and Belgium in banning TikTok on government devices. And FBI director Christopher Wray warned lawmakers that misinformation spread through the app can “divide Americans.” At the start of the month, Senate intelligence chair Mark Warner unveiled a new measure, the Restrict Act, which would enable the US commerce secretary to ban TikTok and any other tech from six “hostile” nations that the US intelligence community deems a national security threat.

The White House supports Warner’s bill, and the multi-agency Committee on Foreign Investment in the United States told TikTok it would be banned unless it’s completely divested from ByteDance. Lawmakers in both parties welcomed that announcement. “It’s a step toward banning them,” says senator Roger Wicker, a Mississippi Republican.

While many lawmakers embraced White House pressure, they aren’t sitting on the sidelines as they did when former president Donald Trump tried to ban TikTok through executive orders, which the courts ultimately shot down.

“I think it's absolutely a good first step. I'm not sure it's going to get us everywhere we want to be,” says senator John Hickenlooper, a Colorado Democrat. “I don't think we want to have a Chinese-owned company to have that kind of access to not just our kids, our culture.”

It’s not just espionage that worries lawmakers, who argue the app has a vulnerable captive audience among young people in the US. TikTok recently unveiled new efforts to limit users’ exposure to the app, including an hour-per-day time limit for children under 18, but it fails to address lawmaker’s concerns. “I'm not sure TikTok is a healthy ingredient to add to our children's psychological diet,” Hickenlooper says.

TikTok did not respond to WIRED's request for comment.

More than half of states prohibit TikTok’s use on government devices, and it’s banned by dozens of public schools, from grade schools to some of the nation’s largest universities. Tennessee attorney general Jonathan Skrmetti is leading an investigation on behalf of 46 states into whether the app is detrimental to children’s mental health. The US Department of Justice, meanwhile, is investigating reports that TikTok staffers spied on US journalists.

In short, Chew faces a daunting assignment given the hostility that awaits him at 10 am ET tomorrow. 

The debate at the Capitol is now over _how_ to punish TikTok, not _whether_ to punish TikTok. After supporting the Senate’s unanimous consent agreement to ban the app on government devices in December, senator Rand Paul, a Kentucky Republican, is one of the few lawmakers opposed to an outright ban. “I’m against banning TikTok. I think it violates the First Amendment. I think it also violates the prohibition on bills of attainder where one company is targeted by government,” Paul says. 

Other lawmakers dismiss claims that a TikTok ban would be unconstitutional. “We're not regulating the speech of the company, we're regulating the way it uses its data and its ownership on how it creates a national security vulnerability,” says Marco Rubio, the Senate Intelligence Committee vice chair and Florida Republican. 

With the Senate in Democrats’ hands and the House controlled by Republicans, TikTok is one of the few areas where the two chambers seem to agree. 

“TikTok in its current construct is unacceptable,” says senator Lindsey Graham, a South Carolina Republican. “I'd like to come up with a construct that people could enjoy the service, like 100 million people do, and not empower communist China. That's my preferred outcome.”  

With broader data privacy measures stalled in Congress and antitrust measures shelved by the new House Republican majority, TikTok has become the Big Tech bogeyman for both parties. And both parties seem fine with that—at least for now. “The reason why this is a good place to start is because we’re all agreed,” Krishnamoorthi says. “You may disagree about which American companies handle our data best, but nobody disagrees that the Chinese Communist Party should not have access to our private data.”

The TikTok CEO’s Face-Off With Congress Is Doomed

Doctored images of the former US president went viral on Twitter. These are the telltale signs that they aren’t what they seem.

The viral, AI-generated images of Donald Trump’s arrest you may be seeing on social media are definitely fake. But some of these photorealistic creations are pretty convincing. Others look more like stills from a video game or a lucid dream. A Twitter thread by Eliot Higgins, a founder of Bellingcat, that shows Trump getting swarmed by synthetic cops, running around on the lam, and picking out a prison jumpsuit was viewed over 3 million times on the social media platform.

What does Higgins think viewers can do to tell the difference between fake, AI images, like the ones in his post, from real photographs that may come out of the former president’s potential arrest? 

“Having created a lot of images for the thread, it's apparent that it often **focuses on the first object** described—in this case, the various Trump family members—with everything around it often having more flaws,” Higgins said over email. Look outside of the image’s focal point. Does the rest of the image appear to be an afterthought?

Even though the newest versions of AI-image tools, like Midjourney (version 5 of which was used for the aforementioned thread) and Stable Diffusion, are making considerable progress, mistakes in the smaller details remain a common sign of fake images. As AI art grows in popularity, many artists point out that the algorithms still struggle to replicate the human body in a consistent, natural manner. 

Looking at the AI images of Trump from the Twitter thread, the face looks fairly convincing in many of the posts, as do the hands, but his **body proportions may look contorted** or melted into a nearby police officer. Even though it’s obvious now, it’s possible that the algorithm might be able to avoid peculiar-looking body parts with more training and refinement.

Need another tell? Look for odd writing on the walls, clothing, or other visible items. Higgins points toward **messy text** as a way to differentiate fake images from real photos. For example, the police wear badges, hats, and other documents that appear to have lettering, at first glance, in the fake images of officers arresting Trump. Upon closer inspection, the words are nonsensical.

An additional way you can sometimes tell an image is generated by AI is by noticing **over-the-top facial expressions**. “I've also noticed that if you ask for expressions, Midjourney tends to render them in an exaggerated way, with skin creases from things like smiling being very pronounced,” Higgins said. The pained expression on Melania Trump’s face looks more like a re-creation of Edvard Munch’s _The Scream_ or a still from some unreleased A24 horror movie than a snapshot from a human photographer.

Keep in mind that world leaders, celebrities, social media influencers, and anyone with large quantities of photos circulating online may appear more convincing in deepfaked photos than AI-generated images of people with less of a visible internet presence. “It's clear that the more famous a person is, the more images the AI has had to learn from,” Higgins said. “So very **famous people are rendered extremely well**, while less famous people are usually a bit wonky.” For more peace of mind about the algorithm’s ability to re-create your face, it might be worth thinking twice before posting a photo dump of selfies after a fun night out with friends. (Though it’s likely that the AI generators have already scraped your image data from the web.)

In the lead-up to the next US presidential election, what is Twitter’s policy about AI-generated images? The social media platform’s current policy reads, in part, “You may not share synthetic, manipulated, or out-of-context media that may deceive or confuse people and lead to harm (‘misleading media’).” Twitter carves out multiple exceptions for memes, commentary, and posts not created with the intention to mislead viewers. 

Just a few years ago, it was almost unfathomable that the average person would soon be able to fabricate photorealistic deepfakes of world leaders at home. As AI images become harder to differentiate from the real deal, social media platforms may need to reevaluate their approach to synthetic content and attempt to find ways of guiding users through the complex and often unsettling world of generative AI.

How You Can Tell the AI Images of Trump’s Arrest Are Deepfakes

Amid ongoing protests, the Iranian regime has lost control of its image, pushing it to employ increasingly drastic tactics where everyone loses.

In the early hours of January 5, a popular anonymous Iranian dissident account called Jupiter announced on Twitter that his friends had killed Abolqasem Salavati, a maligned magistrate nicknamed the “Judge of Death.” The tweet went viral, and thousands of jubilant people poured into the account’s Twitter Space to thank them for assassinating the man responsible for sentencing hundreds of political prisoners to die.

Soon, however, a few attendees voiced doubts over the veracity of the claim. They were cursed at and kicked out of the room, as the host insisted, “Tonight is about celebration!” while repeatedly encouraging viewers to make the Space go viral. The next day, activists on the ground and Iranian media confirmed that Salavati was, in fact, alive. Several experts suspect Jupiter to have been an Islamic Republic of Iran cyber operation aimed at distracting people, while the Iranian government executed two protesters the same night as the Twitter Space.

Within its borders, the Iranian regime controls its population through one of the world’s toughest internet filtering systems, physical crackdowns, and mass arrests carried out with impunity. However, the IRI is vulnerable beyond its physical and virtual borders, as the regime struggles to contain the discourse and silence dissidents. To combat opposition narratives in the West and among VPN-armed domestic activists online, the IRI cyber army deploys multifaceted, devious, and sometimes clumsy tactics. With the ongoing political unrest in Iran, old cyber tactics have been ramped up, and new tricks that aim to distract, discredit, distort, and sow distrust have come to the fore as the regime finds itself in a critical moment.

Desperate Times, Desperate Measures

Among the tactics used by the IRI’s cyber agents—known colloquially as Cyberi—is old-school hacking. The Iran-linked hacker group Charming Kitten gained notoriety in 2020 for its spear-phishing attempts on journalists, scholars, and policy experts in the West. The group was recognized by its signature strategy of pretending to be reporters or researchers and feigning interest in their targets’ work as a pretext for setting up interview requests embedded with a spear-phishing link. Recent reports from the UK government’s National Cyber Security Center and security firm Mandiant found that such spear-phishing activities cyber groups TA453 and APT42, which are affiliated with the Iranian Revolutionary Guard Corps, have been increasingly prevalent. Last month, the popular anti-regime account RKOT claimed to have received an interview request geolocated to an IRGC department in Shiraz from an individual purporting to be a journalist from _The New York Times_. 

According to Amin Sabeti, founder of CERTFA, a cybersecurity collective specializing in uncovering state-backed Iranian cyber activities, these operations have shifted their methods over the past few months, since most targets of interest are aware of the threat and have learned to protect themselves from spear-phishing. Instead, Sabeti says, they now use a “domino effect” strategy by taking aim at low-profile targets, whose credentials they harvest in order to build trust and gain access to higher-profile targets in their network. Early this month, for example, the Iranian Canadian human rights activist Nazanin Afshin Jam said that she received a spear-phishing link from a trusted colleague who had been hacked.

“Right now, they go after everyone who they are interested in, in terms of this revolution, especially people who are working in nonprofits,” Sabeti says. 

Notably, some of these state actors establish credibility and trust over time by masking themselves as anti-regime voices and ardent supporters of the protest movement, or by building relationships with targets. One account by the name of Sara Shokouhi was created in October 2022 and claimed to be a Middle East scholar. The account spent months boosting opposition voices and writing heartfelt tributes to protesters before finally being outed by Iran experts as a state-sponsored phishing operation.

The ongoing protest movement in Iran has also given rise to more intense disinformation campaigns on social media. It is difficult to attribute the activities of anonymous accounts to the Iranian regime with certainty, since many regular Iranians are anonymous for security reasons. Occasionally, however, sloppiness—failing to delete old tweets or accidentally posting pro-regime messages on opposition sock puppet accounts—have revealed their likely affiliation with IRI intelligence.

The fast-paced, chaotic environment during protests within Iran has been ripe for disinformation and information pollution, making it too overwhelming to verify every claim even for researchers, says Simin Kargar, a nonresident fellow at the Atlantic Council’s Digital Forensic Research Lab. 

According to Kargar, the prospects of IRI disinformation operations shifting the narrative in the regime’s favor are slim. However, its cyber activity has had the effect of sowing doubt, where no one is sure what is true and who is trustworthy.

“Iranian information operations haven’t been successful in terms of the region and the type of attention they tend to draw,” she says. “But if their sole accomplishment in this situation is to make everyone doubt whatever is said, even by credible people, that's a big accomplishment for them. It's about creating a climate of fear and intimidation more than making people believe them. Because at this point of 44 years of the Islamic Republic, no one believes what they say.”

No One to Trust

Although widespread vigilance against IRI state actors has made the public less susceptible to propaganda tactics, it has simultaneously created an environment of distrust, where anybody is potentially a regime goon.

A recent study by Hossein Kermani, a political communications researcher at the University of Vienna, outlines the various tactics deployed by the Iranian cyber army that muddy the waters: “Downgrading discussions to the level of the government, justifying the state’s policies, cheering up other users, portraying that everything is normal, redirecting debates, spreading fake news, trending misleading hashtags, and mocking dissidents and activists.”

Some of these strategies were used to accuse the two journalists who covered the death of Mahsa Amini of being Mossad agents, portraying the prominent Iran-based activist Sepideh Qolian as “mentally ill and hysterical,” and pretending to be activists calling for violent uprisings in order to justify violent crackdowns. Pro-regime accounts also mimicked and hijacked anti-regime hashtags, which were subsequently flooded with pro-regime misinformation. And IRI cyber forces boosted the “Do execute” hashtag in response to activists trending “Do not execute” against the string of death penalty verdicts against protesters.

Beyond causing confusion, discrediting and undermining opposition has been an essential component of Iranian cyber activity. This has partially been pursued by hacking opposition figures directly but also through sock puppet accounts, which pit one faction of the opposition against another.

Kargar, of the Atlantic Council, believes that the Iranian cyber response to the recent anti-regime movement signals a smarter and more prudent approach. She describes how regime-affiliated cyber groups hack as much as possible and hold onto their material, sometimes for years, only to release it when it can have maximum effect. These efforts are part of a wider effort to create the impression that nobody is trustworthy and nobody is credible, she says. 

In late 2022, for instance, actress and activist Nazanin Boniadi began publicly collaborating with the son of the former shah, Reza Pahlavi, another popular opposition figure. Amid their attempts to present a unified front against the regime, the IRGC-affiliated hacker group Adl-e Ali sought to sow division by leaking old emails of Boniadi criticizing Pahlavi’s base of staunch monarchist supporters. With Pahlavi seeing a surge of popularity and credibility in the eyes of Western media in recent months, the hacker group has attempted to embarrass and discredit him by leaking old private photos and videos of Pahlavi snoring in his sleep, lying on the beach, and attending parties.

“If you can't win the hearts and minds, you might as well try to make everyone lose hearts and minds in one another. If you can’t win—make everyone else lose. That’s basically how they have been operating,” Kargar says.

While the regime might have lost the battle for the narrative, it has still had some successes, such as the “Judge of Death” distraction. Sabeti, the cybersecurity expert, believes that the case was an attempt by IRI cyber forces to test the waters toward future activities.

“I think it was to maybe the test the waters, to see how they can manage to change the narrative for the next \[execution\],” Sabeti says. “Because everyone knows it’s not the end of the story.”

Source

Wired