Brazil's Fake News Legislation Moves Forward, Gets Slightly Better And Way Worse

from the land-of-contrasts-with-some-contrasts-more-equal-than-others dept

Taking a cue from the then Demagogue-in-Chief of the United States, Brazil's government decided something must be done to control the spread of "fake news" to give the government more control of the narrative. "Fake news" continues to be a handy concept to abuse by governments seeking to limit their constituents' ability to consume or create content. That was Donald Trump's rationale as well, even if it was never articulated with any clarity or cohesiveness.

Brazil's government liked the abusability of "fake news" and turned it into legislation. But before it did that, it gave itself some more power. It "mobilized" the federal police to combat "fake news" ahead of the country's 2018 elections. It gave people with guns the power to regulate the internet, allowing them to "identify and punish" people who published whatever those in power said was "fake."

The proposed law made a bad thing even worse. The original proposal demanded that social networks and messaging apps tie accounts to users via the country's national identity cards. It also mandated data collection on those who spread "fake news" to determine how far it had carried and, if possible, discover the origin of the content. Since there's no way to tell what the government would call "fake news" and when the government might get around to making this determination, the safest way for services to comply would be to log all info indefinitely just in case the government decided it needed access to this info later.

The bill has since been revised. And while it's not quite the privacy nightmare it was when it was introduced, it really hasn't improved much. The EFF has two posts covering the legislation, detailing what's been made better, what's been made worse, and what has remained terrible since the bill's introduction last summer.

Here's the good news:

Regarding serious flaws EFF previously pointed out, the bill no longer sets a general regime for users' legal identification. Second, it does not require social media and messaging companies to provide their staff in Brazil remote access to user logs and databases, a provision that would bypass international cooperation safeguards and create privacy and security risks. Most importantly, it drops the traceability mandate for instant messaging applications, under which forwarding information would be tracked.  

But several downsides remain. First, there's still some mandated unmasking of users, via mandates on publication of user info of those paying for content mentioning political parties as well as those buying or authorizing political advertising. This will make it insanely easy for powerful political officials to target anyone who opposes or criticizes them, adding to list of laws that can be abused to silence their opposition.

The proposal also gives the government the power to pull the plug on parts of the internet if it decides the law has been ignored or violated.

The bill also establishes that internet platforms can have their activities prohibited or temporarily suspended as part of the penalties for noncompliance. According to Article 2, the draft law applies to social networks, search engines, and instant messaging service providers with over two million users registered in Brazil.

Too many critics hanging out in certain internet places? Shut it down. Not only does this provide the government with a quick way to silence up to two million people at a time, but it directly contravenes UN agreements that forbid the wholesale disruption of communication services relied on by countries' populations.

Back to improvements. The amended bill finally places some limits on data retention by platforms and communications apps, restricting it to communications metadata. And that collection is further limited by time constraints: only 15 days in a row, extendable to a maximum of 60 days via court approval.

But, even these improvements have some serious issues. In the EFF's second post, Veridiana Alimonti points out the government thinks it can tie an IP address to a person, which will allow it to make even worse assumptions about communications metadata that go beyond its original faulty assumption: that an IP address = a person.

The provision forces internet applications to unequivocally individualize the user of an IP address, apparently based on the flawed aspiration of linking a given IP address to a specific user without a margin of error. This language offers wide-open interpretations by law enforcement and courts that could severely extend the current data retention mandates, or even force the use of persistent identifiers linked to our every single move online. There are so many variables in internet routing that it is not possible for an application to say unequivocally who is related to a connection.

That assumption was faulty back when most internet communication was handled via desktop computers at people's homes. Mobile devices, public Wi-Fi, limited availability of IPv4 addresses, and the use of multiple connected devices in any home make this bad assumption completely unworkable. And yet, legislators appear to believe it will work.

If it doesn't work, it can fine and/or pull the plug on service providers that fail to do the impossible. The only way to even attempt to comply is to log all sorts of information that can possibly tie users to IP addresses and make this all available to the government when it's requested. This essentially makes the metadata-only demands meaningless since the government clearly wants to be able to identify users and will expect the same providers that thought they were only obligated to log certain things to collect it all and let the government sort it out.

Filed Under: brazil, censorship, fake news, internet regulations

Chinese Internet Companies Are Censoring People Who Write Or Speak Tibetan Or Uyghur, Lending A Hand To China's Cultural Genocides

from the remembering-Jack-Ma dept

Techdirt has reported on the oppression of Tibetans by the Chinese authorities for 15 years now. More recently, the Turkic-speaking Uyghurs in Xinjiang have come in for the same treatment, with the apparent aim of breaking their spirit and imposing total obedience. But alongside the hundreds of prisons and physical repression -- sometimes leading to deaths -- the Chinese authorities have been making it increasingly hard for Tibetans and Uyghurs to preserve their distinctive, non-Han cultures. Now Chinese Internet companies are lending a hand to these cultural genocides, reported here by Protocol:

First it was Talkmate, a language-learning app that partners with UNESCO, that posted via its official Weibo account that it had "temporarily" taken down Tibetan and Uyghur language classes "due to government policies." There is no set date for them to return.

On some services, even people who already speak those languages aren't allowed to write them. The popular Chinese streaming service Bilibili has banned comments posted in Tibetan and Uyghur:

Screen recordings shared by Fergus Ryan, a senior analyst with ASPI's International Cyber Policy Centre, showed that when he tried to type comments in Uyghur and Tibetan, he received error messages that read: "Comment contains sensitive information."

Similarly, on Douyin, the original Chinese version of TikTok, whenever live-streamers speak an ethnic minority language or a dialect, they will receive a warning to switch to Standard Chinese. And if they don't, Douyin's moderators will just cut off the stream, regardless of the content.

The companies are probably not doing this with the explicit intent to stifle Tibetan and Uyghur cultures. It is more likely that they are frightened they will be punished if they let any content that the Chinese government might deem to be "instigating" terrorism or separatism slip through. The dramatic fall from grace of China's outspoken tech billionaire, Jack Ma, stands as a chilling warning to all Internet companies, big or small. Safer just to block everything in these sensitive languages, it seems. And so the rich living cultures of Tibet and the Uyghurs move another step closer to extinction.

Follow me @glynmoody on Twitter, Diaspora, or Mastodon.

Filed Under: censorship, china, culture, internet, tibet, uyghurs, xinjiang

Gasp! YouTube Shutters Account For Person Committing DMCA Takedown Fraud!

from the praise-be dept

It should come as no surprise to regular Techdirt readers that the DMCA takedown process is not only wide open to fraud and abuse, but that those avenues are regularly used in real life for just those purposes. Takedowns to silence criticism, takedowns to try to steal traffic from others, or takedowns or monetization claims just to get some paltry amounts of streaming revenue: the point is that this shit happens all the time. What absolutely does not happen with any frequency is the folks behind these bogus actions getting punished in any way for their behavior.

And so when it does happen, it's worth highlighting it, putting it on a podium, and trying to have it become the norm rather than the exception. To that end, let's highlight YouTube actually banning an account that appeared to be sending fraudulent DMCA notices.

Yesterday, the owner of the “Musical Creator” channel complained that their channel had been terminated after submitting a copyright takedown request. The operator of the channel doesn’t appear to be a native English speaker as the comment below shows, but it’s clear that they disagree with this decision.

“My youtube channel MUSICAL CREATOR has been terminated due to inlegal information copyright infringement. But I not violating any policy of youtube, I don’t know how my channel is terminated I filled out all the information legal. I want to get back my channel again,” the operator writes.

Now, a couple of things to say here. First, the channel's name is "MUSICAL CREATOR"? Hmm, okay, totally doesn't sound like the kind of thing you would create just to send DMCA notices and try to monetize others' content. And we don't have any details from YouTube as to what specifically was the issue with whatever takedown notice MUSICAL CREATOR sent. Instead, all we have is this from YouTube's notice directly.

“We are concerned that some of the information within this legal request may be fraudulent,” YouTube’s response reads. “Please understand that YouTube receives a large number of fraudulent copyright takedown requests, and we take abuse of that process very seriously.”

While that doesn't tell us much or let us vet out whether fraud was actually committed or not, it's worth keeping in mind again that YouTube takes very little action on this sort of thing generally speaking. By which I mean that the fact it took action in this case leads me to believe YouTube thinks it's on very, very solid footing here.

As a result, the MUSICAL CREATOR account is down. Anyone going to its page is simply told that the account is offline due to "a violation of YouTube's Terms of Service."

Again, it's hard to be completely assured that YouTube's actions in this case are appropriate... but I suspect they are. And, purely as a matter of percentage cases, the fact is that YouTube and other platforms have historically been so vastly on the side of those issuing DMCA notices instead of the potential victims of fraud that it makes stories like this unique.

Even though they shouldn't be.

Filed Under: censorship, copyfraud, copyright, dmca, dmca 512, notice and takedown, takedown fraud, youtube
Companies: youtube

LinkedIn Caves Again, Blocks US Journalists' Accounts In China

from the block-enough-accounts-and-it's-really-not-a-network,-is-it dept

LinkedIn -- the business-oriented social media platform owned by Microsoft -- has spent the last few years increasing its compliance with the Chinese government's demands for censorship. A couple of years back, the network drew heat for not only blocking accounts of Chinese pro-democracy activists but also critics of the government located elsewhere in the world.

The blocking only occurred in China, but that was enough to cause PR trouble for LinkedIn, which restored some of the accounts following some deserved backlash. The Chinese government didn't care much for LinkedIn's temporary capitulations so it turned up the heat. After failing to block enough content, the Chinese government ordered LinkedIn's local office to perform a self-audit and report on its findings to the country's internet regulator. It was also blocked from signing up any new Chinese citizens for 30 days.

The pressure appears to have worked. China is again asking for censorship of voices it doesn't like. And, again, LinkedIn is complying. Here's the report from Bethany Allen-Ebrahimian of Axios, who was one of those targeted by the latest round of account blocking.

LinkedIn blocked the profiles of several U.S. journalists from the company's China-based platform this week, citing "prohibited content." My account was one of the profiles affected.

[...]

LinkedIn customer service sent me an email on Sept. 27 stating that, due to "prohibited content" in the summary section of my profile, the company was blocking my profile from being viewable in China.

• Melissa Chan, a former China correspondent who now works as a journalist in Berlin, posted on Twitter that she had received a similar email on Sept. 28.

• Greg Bruno, the author of a book about China's soft-power push against Tibetans, also posted on Twitter on Sept. 28 that he had received an email from LinkedIn. It cited the "publications" section of his profile, in which the only publication listed is his book.

The Chinese government's ability to block foreign journalists from appearing on the local version of LinkedIn has drawn more attention to its censorship practices. But that's pretty much the thing that China does all the time, so it's hardly surprising it would seek to keep its citizens from interacting with people it doesn't approve of.

What's more disturbing is Microsoft/LinkedIn's compliance, which has drawn the attention of Senator Rick Scott, who is now asking the company to explain why it's helping China with its censorship.

The letter [PDF] from Scott makes some good points, but makes some of those points badly.

The censorship of these journalists raises serious questions about Microsoft’s intentions and its commitment to standing up against Communist China’s horrific human rights abuses and repeated attacks against democracy. These acts of censorship by your company, and the apparent broader Microsoft censorship policy of, “offering a localized version of LinkedIn in China,” is gross appeasement and an act of submission to Communist China.

That's a good point. The follow-up is a bit more self-serving.

While Microsoft is censoring journalists abroad, it is actively spreading misinformation domestically. In March 2021, Microsoft openly decried an election security law passed by the Georgia legislature which made it easier for Georgia residents to vote while reducing the possibility of fraud.

Meanwhile, your company has been silent on Communist China rigging its elections and General Secretary Xi declaring himself ruler for life. In the face of these true assaults on democracy, Microsoft is openly suppressing those who try to expose Xi’s authoritarian rule.

The election security law is really about voter suppression, something Rick Scott certainly isn't going to publicly acknowledge. This is some useless point scoring thrown in to make it appear Microsoft's opposition to this bill is on par with its acquiescence to authoritarian rulers in another part of the world.

Here's what the supposedly pro-democracy "election security" bill passed in Georgia actually does:

In Georgia, Governor Brian Kemp just signed into law a bill that adds many obstacles to voting, including reducing the number of ballot boxes, shrinking the window for early voting, adding additional photo ID requirements, and allowing state officials to circumvent the work of county election officials if they don’t like the outcomes they are seeing. The Georgia bill even goes so far as to make it illegal for outside groups to give water or food to voters stuck in long lines.

It's not about security. It's about making democracy accessible to those with the most resources and expendable time. So, that's some bullshit and Scott would have been better off sticking to chastising LinkedIn for carrying the Chinese government's censorship water rather than pretending his associates in Georgia are all about that democracy.

Scott does ask some good questions, though, which hopefully will generate some explanation for LinkedIn's decision to engage in proxy censorship to maintain access to a sizable number of Chinese users.

Is censorship of user views that are not aligned with the Chinese Communist Party a function of LinkedIn’s “localized version” of its platform in China?

How many accounts has LinkedIn censored because of content the Chinese Communist Party disliked, or that LinkedIn feared might upset Chinese Communist Party authorities?

Why does Microsoft choose to weigh in on domestic political matters, but stay silent on foreign political matters, and what is its decision-making process for speaking on political matters?

The last question is still a good one, even though Scott undercut this point earlier by claiming a voter suppression bill was actually an election security bill. And it could be that LinkedIn does weigh in on local issues in China. It's just that the Chinese government doesn't care what LinkedIn thinks and has offered the platform the option of complying or leaving.

So far, the answer is still compliance. And that's really not an acceptable tradeoff. American companies shouldn't willingly do business with authoritarians. Any compliance only affirms that the party holding the real power is the political party issuing the orders. Demands for censorship will only increase. And opening a local office pretty much guarantees the government will start demanding full access to Chinese residents' data and communications. If you cave on the easier stuff, you've got nothing left to stand on when they come for the rest of it.

Filed Under: censorship, china, content moderation, journalism
Companies: linkedin, microsoft

Copyright Continues To Be Abused To Censor Critics By Entities Both Big And Small

from the tools-of-abuse dept

We've talked far too many times about how the DMCA takedown processes across internet industries as they stand are wide, wide open for abuse. From churches wielding copyright to attempt to silence critics engaging in protected speech, to lawyers using copyright to try to silence critics engaging in protected speech, to freaking political candidates abusing YouTube's DMCA notice process to silence critics engaging in protected speech... well, you get the idea. The point is that we've known for a long, long time that the current method by which the country and companies currently enforce copyright law tilts so heavily towards the accuser that it's an obvious avenue for misuse.

And this is an issue created by bad actors big and small. Hell, apparently you cannot even criticique a sophomoric prank joke troop on YouTube without being targeted using copyright law.

Last week, Tripping, a smallish YouTube creator with about 88,000 subscribers, faced an uphill battle to keep one of his videos up. It was pretty standard as far as social commentary videos go on YouTube: an 11-minute presentation about “The Rise and Fall of NELK,” a prank channel that has nearly 7 million subscribers.

But since Tripping posted it in May, it’s been flagged for “copyright” infringement several times. Last week, it was successfully taken down by YouTube before being reinstated over the weekend, when 21-year-old Quentin, who owns the account, disputed it. But Quentin is now concerned that his video criticizing the pranksters was successfully censored by the Nelk Boys themselves through a loophole he says YouTube isn’t vetting carefully enough.

That appears to be precisely what happened. The back and forth over this 11 minute video is fairly amazing. The whole video called out the Nelk Boys for doing a bunch of dumb stuff, most specifically with regards to ignoring COVID and encouraging others to do the same. Other behavior was questioned as well. A week after posting it, Quentin's video got flagged for a copyright claim and was blocked. He fought that claim and the video was put back online. Then, months later, Quentin was assessed his first copyright strike by YouTube, meaning his account was now in jeapordy of being shut down if future strikes occurred.

And the email from YouTube listed Kyle Forgeard of the Nelk Boys as the copyright owner issuing the claim resulting in the copyright strike.

“They’re using copyright ... and YouTube’s system as a way to avoid the criticism; it’s a very common thing on YouTube,” he said. “There’s no way for YouTube to determine what’s this or that. If Nelk [claims copyright], they’re going to have to accept it out of respect for Nelk and YouTube not getting sued. But with fair use and for commentary, you’re allowed to use clips in that manner.”

When I reached out, a spokesperson at YouTube said users can file a dispute if they think their account was flagged erroneously, and, in this case, it was. I then asked them how YouTube is going to improve this system — especially when algorithmic moderation has become a widespread issue time and time again — but I did not hear back.

And that's why nothing ever changes. The onus for getting protected speech back online, in the vaunted opinion of YouTube, falls squarely on the victim of the censorship-by-copyright tactic. Any time you try to kickstart a conversation about how to improve this system to make it balanced instead of tilted towards the claimaint, you get crickets. And the end result is the YouTube creator community being completely rudderless as to what exactly they're supposed to do.

Quentin said he feels somewhat powerless when something like this comes up: “[Nelk] are so much larger than me. They have millions of subscribers and fans, and I only have a minimal fraction of what they have. All I’m betting on is YouTube is finally reviewing it and coming to terms with what’s right.”

He added, “I wouldn’t be surprised if they still want that video down.”

He just hopes YouTube “implement[s] a system that prevents larger channels from filing claims like this and have it go directly through because copyright is a serious thing.”

As is fraud and censorship, I might add. But for far too long YouTube has been reluctant to do anything about that side of this equation, preferring instead to molify the copyright industries and leave some segment of its user base hung out to dry.

Filed Under: censorship, commentary, copyright, dmca, takedowns, tripping
Companies: youtube

Texas' Unconstitutional Social Media Censorship Bill Challenged In Court, Just As Texas Joins The Legal Fight For Florida's Unconstitutional Social Media Bill

from the florida-and-texas;-what-a-pair dept

Texas and Florida. Florida and Texas. Two states with governors who have decided that culture warrioring and "owning the libs" is way more important than the Constitution they swore to protect and uphold. As you'll recall, last month Texas Governor Greg Abbott decided to use the internet services he hates to livestream his signing of the clearly unconstitutional HB20 that seeks to block social media sites from moderating how they see fit.

As we had pointed out, Florida had beaten Texas to the punch on that and a court had already tossed out the bill as an unconstitutional infringement of 1st Amendment rights. Now a state that was looking to actually do things correctly would maybe see that and recognize that maybe it's not worth wasting millions of taxpayer dollars to do the exact same thing, but Texas went ahead.

And, now, the same two organizations that sued to strike down Florida's law, NetChoice and CCIA, have similarly sued to strike down Texas' law.

At bottom, H.B. 20 imposes impermissible content- and viewpoint-based classifications to compel a select few platforms to publish speech and speakers that violate the platforms’ policies—and to present that speech the same way the platforms present other speech that does not violate their policies. Furthermore, H.B. 20 prohibits the platforms from engaging in their own expression to label or comment on the expression they are now compelled to disseminate. And in light of the statute’s vague operating provisions, every single editorial and operational choice platforms make could subject those companies to myriad lawsuits.

These restrictions—by striking at the heart of protected expression and editorial judgment—will prohibit platforms from taking action to protect themselves, their users, advertisers, and the public more generally from harmful and objectionable matter. At a minimum, H.B. 20 would unconstitutionally require platforms like YouTube and Facebook to disseminate, for example, pro-Nazi speech, terrorist propaganda, foreign government disinformation, and medical misinformation. In fact, legislators rejected amendments that would explicitly allow platforms to exclude vaccine misinformation, terrorist content, and Holocaust denial.

Additional H.B. 20 provisions will work to chill the exercise of platforms’ First Amendment rights to exercise their own editorial discretion and to be free from state-compelled speech. H.B. 20 will impose operational mandates and disclosure requirements designed to prescriptively manage—and therefore interfere with and chill—platforms’ exercise of editorial discretion. In a series of intrusive provisions, H.B. 20 requires “social media platforms” to publish how they intend to exercise their discretion, document in excruciating detail how they exercise their editorial discretion over potentially billions of pieces of content, and operate inherently burdensome and unworkable individualized complaint mechanisms—all of which together work to compel or otherwise challenge the platforms’ countless daily uses of editorial discretion.

Notably, the lawsuit does not challenge the email filter provisions in the law, which effectively means that on December 2nd, if no one else tries to stop it, spam filters may be in violation of Texas' law. As Prof. Eric Goldman has noted, any spammer whose email is caught in a spam filter will then be able to sue the filter provider and seek statutory damages. Fun stuff!

Meanwhile, also this week, just to show how totally committed Texas Attorney General Ken Paxton is to unconstitutional restrictions on the free exercise of editorial discretion, he also filed an amicus brief in the appeal of the Florida ruling. A bunch of other states (Alabama, Alaska, Arizona, Arkansas, Kentucky, Mississippi, Missouri, Montana, and South Carolina) all signed on, but this is a Texas product. Also, it's hot garbage. It insists that these bills don't regulate speech -- when they very, very clearly do.

But the district court’s First Amendment analysis is riddled with errors. It veered off course from the outset by concluding that S.B. 7072 regulates speech, when that law instead regulates conduct that is unprotected by the First Amendment: social media platforms’ arbitrary application of their content moderation policies.

I mean, what? Of course, content moderation policies are protected by the 1st Amendment. It's basic editorial discretion.

The entire amicus brief tries to claim that editorial discretion is "conduct" and not speech -- and that would upend basically all 1st Amendment precedent. And if Texas actually got its way with this, then that would enable the government to regulate who could appear on Fox News and other media organizations, claiming that those demands are "conduct, not speech."

Nothing in S.B. 7072’s neutrality and disclosure provisions regulates the speech of Plaintiffs or the members of their trade associations—they “neither limit[] what [Plaintiffs or their members] may say nor require[] them to say anything.” FAIR, 547 U.S. at 60. Instead, at most these provisions regulate the conduct of Plaintiffs and their members: their arbitrary and blunderbuss content-moderation policies

This is... wrong. Requiring them not to moderate certain content (as the Florida bill does for political speech) is absolutely requiring them to associate with speech they may disagree with -- and such compelled association is a violation of the 1st Amendment. Did Ken Paxton actually graduate law school without learning this?

Like Florida did in its case, the Texas Amicus brief relies heavily on Rumsfeld v. FAIR. The district court in Florida rightly pointed out that FAIR does not apply here and is easily distinguished:

The Florida statutes now at issue, unlike the state actions in FAIR and PruneYard, explicitly forbid social media platforms from appending their own statements to posts by some users. And the statutes compel the platforms to change their own speech in other respects, including, for example, by dictating how the platforms may arrange speech on their sites. This is a far greater burden on the platforms’ own speech than was involved in FAIR or PruneYard.

But, Texas insists otherwise -- and now we have to hope that the panel of Judges on the 11th Circuit recognizes the absolute garbage that is Texas' brief.

Filed Under: censorship, content moderation, florida, social media, texas
Companies: ccia, netchoice