Twitter Makes A Bet On Protocols Over Platforms

from the well-this-is-nice dept

It looks like Twitter is making a bet on protocols over platforms for its future.

Nearly five years ago, I first wrote about the idea of why protocols are a better approach than platforms for various internet platforms struggling with content moderation issues. In that first post discussing it, I talked about two companies that I thought might benefit most from such an approach: Twitter and Reddit. Over the intervening years, I've been thinking, talking, and writing on this subject quite a bit, including the big paper I released with the Knight First Amendment Institute at Columbia University, Protocols, Not Platforms: A Technological Approach to Free Speech. That details a variety of different ideas regarding how we might move to a world dominated by protocols, rather than just centralized, proprietary, siloed platforms.

In the paper I talk about whether or not it makes sense for a larger company to adopt a protocol approach and to move away from a platform-based world. I've been talking this over with a variety of folks in and out of the tech world for years, and kept getting told that there was no way any company would voluntarily move in such a direction. And, after watching Facebook's weird approach with Libra (which is already looking like a flop), it only seemed to confirm that many of these companies are simply too entrenched in the platform model.

And yet... just a little while ago, Twitter co-founder and CEO Jack Dorsey has announced that Twitter is going to start experimenting with protocols as a different approach to how the company might architect its business (and, yes, he cites my paper in his announcement). This is not, by any means, them jumping off the high dive into the world of protocols, but more a toe in the water. Basically, Twitter is going to seed fund an outside team, called BlueSky, that will be charged with creating an open and decentralized protocol standard for social media which Twitter might eventually use for its own system.

There are a lot of challenges here -- and many reasons why this project could fail. There are a bunch of existing teams trying to build their own protocol approaches for this as well. But, I think it is notable that Twitter, specifically (the same company I first wrote about regarding this approach five years ago) has decided to take a serious look at moving in this direction. I appreciate that Twitter is not shying away from the challenges and potential pitfalls to this approach, but going in with eyes wide open, suggesting that this is an experiment worth exploring.

If you're worried about the dominance of certain social media platforms, or if you're concerned about privacy online, or if you're uncomfortable with leaving the decisions for how content moderation works in the hands of a few internet company bosses -- this is big news and something you should be paying attention to. It won't change the way the web works overnight. Indeed, it might never have that big of an impact. But it certainly has the potential to be one of the most significant directional shifts for the mainstream internet in decades. Keep watching.

Filed Under: distributed, jack dorsey, platforms, protocols, social media
Companies: bluesky, twitter

Court Says It's Not Tortious Interference To Report Your Account Or Ask Twitter To Ban You

from the 230-wins-again dept

Laura Loomer, professional troll, these days seems most well known for all the internet platforms that have banned her:

Loomer is banned from Facebook, Twitter, Instagram, Medium, Paypal, Venmo, GoFundMe, Uber and Lyft for violating their terms of service, in many cases by using hateful and Islamophobic speech.

That is a pretty impressive list right there. She's apparently trying to run for Congress to use that as an argument for why she should be let back onto these platforms, but so far that's not working either. She's also been suing companies for kicking her off the platform, which seems like a good way to get many lessons in how Section 230 of the Communications Decency Act works. In one such lawsuit, she also sued CAIR, the Council on American-Islamic Relations, arguing that CAIR's "reporting" of her Twitter account to Twitter violated all sorts of laws.

Spoiler alert: it did not. It does not violate any law whatsoever to report an account to Twitter.

The court dismissed a variety of claims for procedural reasons, leaving just a claim of "tortious interference of a business relationship," but that has also been dismissed for a bunch of reasons. First: the lack of any actual business relationship that CAIR could have possibly interfered with. Loomer and her lawyer argued that there was a business relationship between Loomer and her followers, and, separately, between Loomer and Twitter, but the Court was quick to issue a big fat "nope" to both ideas. First, the "business relationship" between Loomer and her followers:

With respect to the relationship between Loomer and her Twitter followers, the Court finds that Loomer has failed to establish an actual business relationship with identifiable customers.... As the Florida Supreme Court has made clear time and again, an action for tortious interference with a business relationship does not lie where the business relationship alleged is one with the “community at large” as opposed to one with identifiable customers.... Here, Plaintiffs fail to identify discernable customers and instead allege a business relationship with Loomer’s Twitter community at large.

Specifically, Loomer contends that Defendant tortiously interfered with the business relationship between Loomer and her Twitter followers by convincing Twitter to ban Loomer from its platform. While the Court accepts all well-pleaded allegations as true, Loomer’s relationship with her Twitter followers—no matter how economically beneficial such a relationship may have been—is a relationship with the community at large, not with identifiable customers.... Given that Loomer’s alleged business relationship with her followers does not sufficiently identify customers, Plaintiffs fail to satisfy the first prong of a claim for tortious interference with a business relationship.

What about her "business relationship" with Twitter? That, too, does not actually exist:

Similarly, with respect to Loomer’s alleged business relationship with Twitter, Plaintiffs have pleaded nothing to suggest Loomer and Twitter had a business relationship as “evidenced by an actual and identifiable understanding or agreement which in all probability would have been completed if the defendant had not interfered.”... For starters, Count II of Plaintiffs’ Amended Complaint is devoid of any allegations suggesting the existence of a business relationship between Loomer and Twitter. Still, at the Hearing, Plaintiffs’ counsel argued that Defendant interfered with Loomer’s right to use Twitter as a platform. However, when pressed to identify the nature of the business relationship between Loomer and Twitter, and Plaintiffs’ corresponding legal rights pursuant to that relationship, Plaintiffs could not do so. Put simply, no business relationship can be inferred merely from Loomer’s decision to use Twitter, and Defendant cannot be found liable for tortious interference with a business relationship where no business relationship has been established.

The court then goes even further and points out that even if Loomer could demonstrate a business relationship with Twitter (which she could not), it wouldn't matter anyway, thanks to CDA 230:

Moreover, even if Plaintiffs had established a business relationship between Loomer and Twitter, Section 230 of the Communications Decency Act (“CDA”), 47 U.S.C. § 230(b)(1)-(2), insulates internet providers, such as Twitter, from exactly the kind of liability Loomer now seeks to impose upon Defendant. Under the CDA, Twitter cannot be held liable for its decision to exercise traditional editorial functions, such as moderating content on its platform. Mezey v. Twitter, Inc., No. 1:18-CV-21069-KMM, 2018 WL 5306769, at *2 (S.D. Fla. July 19, 2018). Here, Plaintiffs seek to hold Defendant liable for just that.

At the Hearing, Plaintiffs’ counsel argued that by urging Twitter to ban Loomer, Defendant had interfered in Loomer and Twitter’s business relationship. However, even accepting as true Plaintiffs’ proposition that Defendant reported Loomer’s account and convinced Twitter to ban Loomer, doing so does not create a cause of action for tortious interference with a business relationship. Plaintiffs’ suggestion that the mere reporting of a Twitter user—however insistent such reporting may be—is sufficient to constitute tortious interference in a business relationship between Twitter and the targeted user is, to put it mildly, nonsensical. A cause of action for tortious interference simply cannot exist each time an individual reports another Twitter user. And at the Hearing, Plaintiffs’ counsel could not overcome this line-drawing problem. Put simply, Plaintiffs cannot use a claim of tortious interference with a business relationship as an end-run around the CDA’s explicit prohibition that Twitter cannot be held liable for its decision to moderate content on its platform.

Why, yes, the argument is nonsensical, and that's probably why the court needed just six quick pages to dismiss the lawsuit.

Filed Under: cda 230, content moderation, laura loomer, section 230, tortious interference
Companies: cair, twitter

Twitter And Instagram Both Begin Experiments In Decreasing The More Socially Questionable Incentives Of Their Platforms

from the fascinating-to-see dept

Two separate news reports last week highlighted how both Twitter and Instagram appear to be taking to heart arguments made about how both of those platforms may (inadvertently) encourage questionable behavior. Instagram will begin hiding "likes" from users in the US to cut down on the dopamine rush of trying to maximize those bits of pointless social validation:

Months after the company tested hiding "like" counts in Australia, Brazil, Canada, Ireland, Italy, Japan, and New Zealand, CEO Adam Mosseri announced today at WIRED25 that some Instagram users in America can expect their like counts to vanish from public view.

The company will begin testing next week, at first rolling out the change to a limited number of accounts.

Meanwhile, Twitter is going to begin experimenting with some small nudges to not rush in to attack people as a first response. The moves here are a lot more subtle than Insgram's hiding of likes and involve encouraging the use of emoji.

In a meeting at its San Francisco headquarters in late October, Gasca and Suzanne Xie, director of product management at Twitter, showed off two experiments among several that will go live in the coming weeks: In the first, Twitter will add an emoji to a retweet, giving people a chance to quote-tweet without going into the compose field. Gasca and Xie want to find out if this feature might encourage people to express more nuanced emotions, putting a damper on dunking and mindless retweeting.

In the second experiment, Twitter will automatically suggest people use an emoji in their replies. If you like something, you could use the heart-eyes emoji. If you don’t, you could use the red circle with a line going through it. But if you pick a negative emoji, Twitter will ask, “Why do you disagree?” — which it hopes will prompt a more thoughtful reply, rather than a flame war.

To be honest, I'm not entirely sure how or why this would work to decrease negative responses -- though at least popping up the interstitial question of "why do you disagree" (as per the second experiment) might drive at least some people to think twice before rushing to dunk on someone.

I'm not sure either of these moves will really change the overall incentives or how these platforms are used, but I find both fascinating for different reasons: a much more clear acknowledgement from both platforms that overall incentives matter, and that small design choices can have outsized influence on how the platforms are used. Given that, it's fascinating to see both platforms then choose to experiment with the little nudges built into the design of their platforms to see how it will play out in terms of usage -- especially regarding socially questionable activities on both platforms.

Of course, I can imagine how their could be some pushback as well, concerning how these are the kinds of subtle paternalistic moves that some people fear will be used to influence behavior in manners that some might not like. I think it's pretty clear that, in both cases above, the sites are aiming to improve the overall "user health" levels on their platform, but I can see how some might (or absolutely will!) claim that they're being used to tamp down on certain viewpoints or ideas -- which, to some extent, highlights why there are no "win" conditions for platforms making these decisions. Every move will be criticized in one way or another.

If you can get past that, however, I think it's a good thing that the platforms are moving to explore how these kinds of tweaks can improve their platforms and how people interact with them.

Filed Under: behavior, emoji, likes, social behavior, social media, social networks, trolling, tweaks
Companies: instagram, twitter

India Is Stifling Kashmir Journalists And Twitter Is Helping Get The Job Done

from the this-isn't-helping dept

India has expressed an interest in being considered a top-tier totalitarian state. Not content to let nearby nations claim all the glory in the "Terrible Places to Live" race, India has been stepping up its censorship and its domestic surveillance, presumably in hopes of being the next country to claim a coveted UN blacklist spot.

The Indian government has asked third party contractors to help it build a massive surveillance network utilizing thousands of cameras and the current cream of the facial recognition crop at the time of deployment. The whole thing needs to be in place less than 8 months after the contract is secured, suggesting the government is more than happy to move forward with whatever it has on hand rather than whatever might actually do the job well.

It's also climbing the global censorship charts, trailing only Russia, China, and Turkey in various social media platform demographics. But it is the king of Facebook censorship, delivering more takedown demands to Facebook than closest rival, Russia. When you're out-censoring Russia, you're playing the censorship game right.

Disputed areas the Indian government claims, but doesn't actually control, are only encouraging further censorship from the Indian government. Unfortunately, US tech companies are helping the government maintain control of the narrative by silencing dissenting voices. Kunal Majumder of the Committee to Protect Journalists has more details:

On August 10, 2018, the Indian government informed Twitter that an account belonging to Kashmir Narrator, a magazine based in Jammu and Kashmir, was breaking Indian law. The magazine had recently published a cover story on a Kashmiri militant who fought against Indian rule. By the end of the month, Indian police had arrested the journalist who wrote it, Aasif Sultan, and Twitter had withheld the magazine’s account in India, blocking local access to more than 5,000 tweets. As of October 2019, Sultan was still in prison, facing terrorism-related charges that CPJ has repeatedly condemned. And the @KashmirNarrator Twitter account was still withheld throughout the country.

There has been almost no improvement over the past year. CPJ reports the government's Kashmir-targeted blackout has only been partially listed. To prevent Kashmir-sympathetic reporting from reaching the territory's residents, the Indian government shut down the region's connection to the outside world by cutting off its internet and phone connections. Some mobile phones have been able to route around the government's blockade, but targeted Twitter accounts remain blocked in the region and journalists have been smuggling out reporting using USB sticks.

Since Twitter is one of the best platforms for spreading news quickly, the Indian government has focused its efforts on this platform. It's following in the footsteps of Turkey and President Recep Erdogan by increasing its targeting of Twitter users. Twitter, unfortunately, has been super-helpful.

A CPJ analysis of those notices reveals that hundreds of thousands of tweets blocked in India since August 2017 under the company’s country withheld content policy were shared by accounts that focus on Kashmir. Among dozens of accounts that were withheld, CPJ identified several besides @KashmirNarrator that were sharing news and opinion, raising serious questions about what safeguards are in place to ensure freedom of the press and the free flow of information.

There are few safeguards built into Indian law. The minimal protections the country allows are not being extended to contested areas the Indian government feels are not deserving of speech protections. Twitter could simply refuse to honor these requests targeting dissenting views, but it has chosen to comply with India's demands, most likely because the country is home to nearly a billion potential users.

If there's any upside to this, it's that the government's targeting appears to be more keyword-based rather than content-based.

CPJ analyzed every account involved, whether the request was based on a single tweet, several tweets, or the account itself, reviewing more than 400 in total. Around 45 percent of those accounts mentioned Kashmir in the handle or bio, or had recently tweeted about Kashmir, according to CPJ’s review.

This censorship is damage that can be routed around. If Kashmir is one of the keywords sought by Indian censors, removing this word might allow links to reporting to sneak past censors. It also appears Twitter is only blocking in certain regions of the country, so tweets and accounts are still visible as long as Twitter believes viewers aren't in the affected areas.

That being said, Twitter's compliance is still problematic. Twitter is unwilling to discuss how it applies local laws to tweets and accounts targeted by governments with track records of stifling dissent. The legal options it has available are generally left unexplored. It makes a certain amount of sense in terms of the bottom line. Challenging foreign rulings and takedown notices requires money and man hours Twitter may not necessarily have to spare. The potential userbase in India is too large to ignore, which means it's probably too large to be considered expendable when principles are on the line.

US tech companies aren't doing all that great dealing with competing interests on a worldwide scale. While some concessions can be excused under the theory that some connection is better than nothing, assisting governments in censorship of critics and journalists isn't acceptable, no matter how much market share is at stake.

Filed Under: censorship, content moderation, india, journalism, kashmir
Companies: twitter

Insider Threats: DOJ Says Twitter Employees Spied On User Accounts For Saudi Arabia

from the internal-controls-matter dept

We live in interesting times. A year ago, the NY Times had reported that the Kingdom of Saudi Arabia was aggressively using Twitter to keep tabs on and harass critics of the government. As part of that story, it also claimed that the Saudis might have a "mole" within the company in the form of Ali Alzabarah, who had risen through the engineering ranks to a point where he could access information on people the Saudi government was interested in. That story only noted that Western intelligence agencies had alerted the company that the Saudis were "grooming" Alzabarah. Now, the DOJ has charged two former Twitter employees, including Alzabarah, along with a third individual who worked in social media marketing, with spying for the Saudis. The complaint is worth reading.

It shows how officials appeared to groom the Twitter employees, starting with Ahmad Abouammo, who was (for a time) a marketing manager at Twitter, but who left the company in 2015. It describes how Saudi officials built up a relationship with him, setting up a tour of Twitter's headquarters, and then later providing gifts, such as an expensive watch. Soon after, Abouammo is accused of accessing information on Saudi critics and dissidents;

Within one week of ABOUAMMO's meeting in London with Foreign Official-1, ABOUAMMO began access private Twitter user information of interest to Foreign Official-1 and the Saudi Royal Family.

On December 12, 2014, ABOUAMMO accessed Twitter User-1's email address through Twitter's computer systems. Based on the FBI's review of Twitter User-1's public postings, Twitter User-1 was a prominent critic of the Kingdom of Saudi Arabia and the Royal Family with over 1,000,000 Twitter followers. ABOUAMMO accessed Twitter User-1's email address again on January 5, 2015, January 27, 2015, February 4, 2015, February 7, 2015, February 18, 2015, and February 24, 2015.

On or about January 17, 2015, Foreign Official-1 began communicating with ABOUAMMO on ABOUAMMO's personal email in addition to his Twitter email account. On January 17, 2015, Foreign Official-1 emailed ABOUAMMO at ABOUAMMO's personal email and stated only "as we discussed in london..." Foreign Offical-1 attached to the email a document that discussed how Twitter User-1 had violated both Twitter policy and Saudi laws. On January 20, 2015, Foreign Official-1 called ABOUAMMO twice (the calls lasting 19 seconds and 51 seconds, respectively). On January 22, 2015, Foreign Official-1 called ABOUAMMO seven times (no answer, 6 seconds, 5 seconds, 3 seconds, 4 minutes 37 sections, 3 minutes 36 seconds, and 39 seconds, respectively). ABOUAMMO also accessed the email address provided by user @KingSalman the same day. The handle @KingSalman was the Twitter Account for Saudi King Salman bin Abdulaziz Al Saud. Foreign Official-1 called ABOUAMMO on January 23, 24, and twice on January 25, 2015 (1 minute 57 seconds, no answer, 1 minute 57 seconds, and 6 seconds, respectively). On January 27, 2015, ABOUAMMO accessed Twitter User-1's user email address again. The next day, Foreign Official-1 called ABOUAMMO three times (11 seconds, 5 seconds, and no answer).

It goes on like this with the Saudi official sending "reports" about different accounts claiming they violate Twitter's policies -- which had nothing to do with Abouammo's role in marketing. And then comes this:

On February 19, 2015, ABOUAMMO created an online user ID and password to access the Bank Audi Account his close relative opened in Lebanon on February 3, 2015. On February 24, 2015, ABOUAMMO accessed Twitter User-1's email address and phone number. On the same day, a $9,963 wire transfer was sent from the Lebanon Bank Audi Account to ABOUAMMO's Bank of America account...

There are a few more such transactions. Even after Abouammo quits Twitter (to go to Amazon) he apparently continues to receive money, and would reach out to former colleagues at Twitter to try to access the information the Saudis were seeking. Also, according to the DOJ, Abouammo lied to the FBI (not a good idea), then tried to "alter" documents he gave to the FBI (a very, very bad idea). Indeed, the DOJ's complaint suggests that while the FBI was talking to him, he asked to go use his computer without them present, and then presented them with what they believe to be a doctored invoice.

During the interview, ABOUAMMO offered to obtain a copy of the consulting contract from a desktop computer that he claimed was in his bedroom, and requested that the Agents not follow him to the bedroom. A few minutes later, ABOUAMMO emailed an undated invoice to the email address of a Palo Alto, California based FBI Special Agent.

I believe that ABOUAMMO created the invoice in his residence on a computer while Agents were waiting for ABOUAMMO to provide them with the invoice that was discussed. The address on the invoice for services ABOUAMMO said were rendered in 2015 and 2016 was ABOUAMMO's current address, however, public records show that the property was built in 2017 and ABOUAMMO did not purchase it until August 2017, and the metadata properties associated with the file emailed to the FBI indicates that the file was created on the date of the interview (i.e., October 20, 2018)....

Yeah, don't do that.

As for Alzabarah, the DOJ has details of him flying to DC to meet with various Saudi officials, and then upon returning to work getting pretty damn busy getting info on Twitter users critical of the Saudi government.

Within one week of returning to San Francisco, ALZABARAH began to access without authorization private data of Twitter users en masse. Specifically, although ALZABARAH had not access the private user data of any Twitter Accounts since February 24, 2015, starting on May 21, 2015, through November 18, 2015, ALZABARAH accessed without authorization through Twitter's computer systems the Twitter user data of over 6,000 Twitter users, including at least 33 usernames for which Saudi Arabian law enforcement had submitted emergency disclosure requests to Twitter. A Twitter Security Engineer informed the FBI that, although ALZABARAH may have had grandfathered access to view user information through an internal Twitter tool called "Profile Viewer," ALZABARAH had no legitimate business purpose as a Site Reliability Engineer to access user accounts. ALZABARAH's job was to help keep the site up and running, which did not involve accessing individual user accounts.

The DOJ then notes that the Saudi official had email notes detailing the locations of some of the key Twitter users that Alzabarah was looking up for them. Alzabarah also created notes for himself about whether or not he was eligible for reward money the Saudi government was offering for information on certain "terrorists." There are also notes about "secur[ing]" his "future and my family's." Alzabarah later fled the US and apparently resigned from Twitter via email after his flight took off from LA.

This is all kind of fascinating. For all the concerns about outside hacking, insider attacks are still considered a much bigger threat and much harder to detect or prevent. While it does sound like there were some lax controls in place to prevent people who shouldn't have access to certain information from getting to that info, that's not all that unexpected in Silicon Valley (and an area that many companies need to significantly improve in). The complaint does suggest that Alzabarah made "unauthorized" access to this information, even though he was allowed to access a tool. That certainly suggests criminal CFAA charges, and some aspects of that seem problematic. While it does appear that there are other laws Alzabarah broke, arguing it's a CFAA violation to use tools he was granted access to seems like a stretch. And, of course, there are broader questions with Abouammo and why he even had access to any of this data at all.

If anything, hopefully these charges wake up many Silicon Valley companies to the value and importance of better internal controls on how data is made accessible. In the rush to build up companies, insider controls are often one of the last things companies care about (if no one uses a service, who cares, and if a service is growing like gangbusters, then there are many other, more important issues to focus on). But companies need to recognize that when these services are used as broadly as they are, they become an attack vector, and employees are often targeted by governments to try to access information. Of course, it's always going to be impossible to stop all access to information -- hell, even the NSA has had to deal with poor controls and "insider" attempts to access information (see: Snowden, Ed). But it does seem increasingly important for companies to be aware of how valuable their data might be to governments, and do more to protect it.

Filed Under: ahmad abouammo, ali alzabarah, cfaa, controls, employee access, espionage, insider threats, saudi arabia, spying
Companies: twitter

Twitter's Decision To Ban Political Ads Is A Moderation Choice Itself That Likely Will Backfire In Its Own Way

from the because-content-moderation-is-impossible-to-do-well dept

Last week we wrote about Twitter's decision to ban all political ads, most likely in response to watching all the shit being flung at Facebook for its decision to not fact check political ads. We focused on the fact that the "costs" of content moderation can sometimes be so high as to make any related revenue just not worth it. However, in that post we did mention that no matter what, there would be criticism of this decision and follow-on decisions concerning what is, and what is not, a "political" advertisement.

There have been a bunch of good, thoughtful articles about all of this that seem worth highlighting. First up is a piece from Markena Kelly at the Verge, who pointed out that Facebook has already tried to ban political ads, but just in the state of Washington, in response to local laws. And just as we predicted will happen with Twitter, there have been ongoing disputes over what constitutes a political ad:

The first major test case for the new system came with Seattle’s city council elections, which will be wrapping in November. Marijuana entrepreneur Logan Bowers ran for city council on an urbanist platform, but he ended up fighting an uphill battle on platforms. He says confusion around the ban “created an unfair and an unlevel playing field and in many ways it made the situation worse.” High-profile ads were ultimately removed by Facebook, usually after they were reported in the media — but plenty of others skated through.

“Some people had their ads restricted and other people didn’t,” Bowers says, usually according to who knew how to spot the loopholes in the system. “Not everyone’s a lawyer.”

And, as that article notes, even though Facebook did this to try to comply with local state laws, the state is still going after Facebook. So it didn't even help on that front. Instead, it was just fights over whose ads get through:

In April, The Stranger reported that one Seattle City Council candidate, Heidi Wills, was able to run a handful of ads on Facebook while her opponent, Kate Martin, was blocked from running any. The two candidates got into a spat through the Wills campaign’s own comments section on Facebook with Martin pleading, “Could you stop paying to promote your Facebook posts and just play by the rules like the rest of us? It’s getting annoying.”

Wills replied, “I am following all the rules and you are welcome to stop following my campaign on FB.”

Meanwhile, Will Oremus, writing for OneZero points out that Twitter's ban is likely to mostly harm activists and organizers:

There’s something to be said for a tech platform taking its responsibilities to the democratic process seriously. But banning political ads is not as straightforward, nor as obviously correct, as those cheering Dorsey’s announcement seem to think.

The problem is twofold. First, defining which ads count as “political” gets tricky in a hurry. Second, prioritizing commercial speech over political speech is itself a political stance, and not necessarily one that we should want our online communication platforms to take.

He notes all sorts of potential downstream problems:

Presumably, tech companies will still be able to run ads touting their commitment to user privacy, but watchdog groups will be barred from running ads suggesting that we need better privacy regulations. Big corporations will be able to boast about how they treat workers, but unions won’t be able to push for prevailing wage laws or workplace safety laws.

Meanwhile, Cat Zakrzewski, writing over at the Washington Post, speaks to some experts who also note that figuring out what counts as a political ad is very, very difficult, leading to plenty of gaming the system (as Facebook discovered above):

Laura Edelson, a PhD candidate at New York University studying political ads on social media, tells me she found instances in the 2018 election where ads from multiple sitting senators were not marked as political ads by Twitter.

That report notes the general problems with classifying ads at all. And while it's framed as something that Twitter is bad at, the reality is that anyone is likely to be bad at this, since so much of this relies on subjective calls, and no one's going to agree.

“The technical problem of enforcing this ban is the same one as enforcing their disclosure requirements, and if they make less information transparent, it will be harder for third parties like us to monitor if they are actually enforcing this policy,” she added.

Shannon McGregor, an assistant professor at The University of Utah, said she and other researchers detected instances where foreign governments were running ads on Twitter subject to the Foreign Agents Registration Act — yet those ads were not identified in Twitter’s ad database. She said if Twitter was unable to identify and include ads from a foreign government in its library, it raises questions about the site's ability to ensure that no political ads about domestic issues are run on its platform.

Meanwhile, the chair of the Federal Election Commission, who in theory has some purview over political advertising, has weighed in as well with a pretty smart take about how banning all political ads doesn't really make much sense. Instead, it might make much more sense to limit the "microtargeting" of such ads.

Here’s a move that would allow political ads while deterring disinformation campaigns, restoring transparency and protecting the robust marketplace of ideas: Sell political ads, but stop the practice of microtargeting those ads.

“Microtargeting” is the sales practice of limiting the scope of an ad’s distribution to precise sets of people, such as single men between 25 and 35 who live in apartments and “like” the Washington Nationals. But just because microtargeted ads can be a good way to sell deodorant does not make them a safe way to sell candidates. It is easy to single out susceptible groups and direct political misinformation to them with little accountability, because the public at large never sees the ad.

As she notes, this approach would enhance transparency and accountability (since people could better see what political ads were on the platform), enable the ability for more people to call out and debunk disinformation, and (in theory -- perhaps optimistically) push political advertisers to create messages that apply to a broader group of people, rather than narrowly targeting certain groups and "fueling the divisiveness that pulls us apart."

Of course, to some extent this still relies on Twitter being able to decide what is and what is not a political ad. We've argued right here in the past that social media platforms should consider dumping targeted ads altogether, but people always yell at us that such ads are way too valuable to completely dump. Still, bringing things back around to our first post on Twitter's decision: perhaps if the "costs" of continuing to run such ads is so high, they'll find that it makes more sense to dump them entirely.

Filed Under: content moderation, fact checking, political ads, scale
Companies: twitter

Man Sues Twitter For $1 Billion Claiming His Account's Suspension Violated His Right To Worship President Trump As A Demigod

from the suing-Twitter-to-own-the-liberal-construing-of-his-complaint dept

Several stupid lawsuits have been brought against social media companies. Some feature actual lawyers (but mostly from the same law firms) helping clients throw money away on allegations that Twitter and Facebook are at least indirectly responsible for terrorist attacks.

Others also use real lawyers, but lawyers willing to misread precedent to declare large social media platforms "public squares" and advance some very questionable arguments about First Amendment violations.

Then there's everyone else: the kind of people who think being temporarily suspended from a platform is a billion dollar Constitutional violation. (h/t Eric Goldman)

In this lawsuit, Adrian Rangel alleges his brief suspension violated the Constitution harder than it's ever been violated before. Rangel's Twitter account is no longer suspended and it's not because he emerged victorious from this lawsuit. It has already been tossed by the federal court.

Rangel's short-lived lawsuit [PDF] asked for $1 billion in damages for his brief suspension, which he alleges violated his First Amendment right to yell "HANG THEM ALL" in a crowded platform. While we can agree Rangel's heated response to "topics from the mundane to the comical" probably should not have resulted in a suspension, we can also agree Twitter's moderation call did none of the following:

Plaintiff contends that by suspending Plaintiff's account religiouserpico Defendants Twitter Foundation and Vijaya Gadde violated Adrian Rangel's constitutional rights to ( (1) freedom of speech, (2) freedom of expression, (3) freedom of religion, (4) freedom of assembly, (5) freedom against unlawful seizure, (6) due process, (7) substantive due process and (8) equal protection of the United States Constitution.

This despite Rangel's claim on his own Twitter feed that he is a "CONSTITUTIONALIST."

Rangel logically points out his "HANG THEM ALL" tweet was not a threat. It targeted no one and was a hyperbolic expression of Rangel's general exasperation with the status quo. (I'm construing his complaint liberally. Rangel doesn't actually reference the tweet his responded to, but judging the rest of the complaint, it probably had something to do with liberal politicians.)

Now, while you might be familiar with misguided assertions that moderation by private platforms violates First Amendment rights (including freedom of assembly), you're probably unaware that account suspensions also violate religious rights.

Before the Defendants suspended Plaintiff's account, Plaintiff used Twitter to proclaim his religious beliefs to the public of being a Born Again King James Bible Only Christian. Plaintiff included the being Born Again King James Bible Only Christian in his Twitter profile. ln addition, Plaintiff followed and was followed by a number of people on Twitter - one group being people of like-minded religious beliefs.

Tangentially, Plaintiff contends that President Donald J Trump was nothing short of miraculously elected by God into the Presidency; most specifically because of Donald Trump's victory in light of the tremendous media, political and social resistance to his election to the Presidency of the United States. Plaintiff used Twitter to support what Plaintiff contends is Donald J Trump's nothing short of miraculous election to the Presidency. As such, Plaintiff's religious beliefs are intertwined with Plaintiff's support of Donald J Trump as President of the United States of America.

Recourse options are limited for those who feel their cult-like admiration of elected leaders has been harmed by moderation efforts. "Limited" as in "zero." There are no options available to someone who has managed to "intertwine" their belief in God with their worship of a president.

Going from that surprising tangent, Rangel alleges the appeals process provided by Twitter doesn't approach the standards of due process provided by the Constitution. Of course they don't. They never will. Only the government has this obligation, much like the government's monopoly on First Amendment rights violations. The same goes for the Equal Protection clause, which is invoked in Rangel's lawsuit to make a perfectly valid point.

Plaintiff further contends that Defendants Twitter Foundation and Vijaya Gadde have illegally embarked upon an illegal circumvention of the United States Constitution in attempting to impose on United States citizens the legal cultures of foreign countries i.e. India, China, Russia, Germany, United Kingdom etc. Many of these foreign countries were once or still are considered third world countries because of their former or present totalitarian subjugation or colonizing regimes.

Unfortunately, valid points aren't the same thing as cognizable claims and the court has no jurisdiction or duty to prevent Twitter from aligning moderation efforts with foreign laws. I agree with Rangel that Twitter should not be humoring authoritarian regimes by complying with removal notices and/or suspending accounts, but a billion dollar lawsuit claiming a private company violated Constitutional rights isn't the place to make this argument.

Since there's no moving forward with the case, there will be no discussion of Section 230 immunity, which would have seen this case dismissed if Twitter (a "California nonprofit," according to the plaintiff) had needed to file a response. Love it or hate it, social media platforms can moderate as they please without violating Constitutional rights. Understanding this simple concept would save a lot of people time and money.

Filed Under: 1st amendment, account suspension, adrian rangel, content moderation, freedom of religion, freedom of speech, lawsuits, pro se, section 230
Companies: twitter

Sometimes The Cost Of Revenue Is Too High: Twitter Bans Political Ads As Facebook Deals With Ongoing Shitshow

from the the-fun-of-content-moderation dept

There's been a lot of talk in the last few weeks about political ads online, kicked off by Facebook "clarifying" that its fact checking rules for regular advertisements don't apply to political ads, after President Trump's campaign ran some ads that were laughably inaccurate. That kicked off a series of political stunts, including Elizabeth Warren taking out her own misleading ads to call out Facebook (though, as we noted, that whole stunt seemed particularly silly since she had previously complained that Facebook shouldn't be blocking political ads -- when they were her own). The debate rages on with everyone insisting that their viewpoint is correct, and with few acknowledging that there is no good answer.

If you fact check political ads, you will undoubtedly be accused of bias against those whose ads get blocked. And a big part of the problem is not about whether or not something is "factual" but about nitpicking around the semantics of what is and what is not a fact, or in how it's presented. This is why most fact checking operations constantly get called out, since so much is a judgment call. And, because of that, there is a reasonable position that Facebook has staked out that when it comes to politics, it doesn't want to be in the business of judging the veracity of one side or another. Of course, that response is wholly unsatisfying and is easy to spin as "letting politicians lie."

And, unsurprisingly, we're now seeing stunts like the one attempted by political activist Adriel Hampton, who has registered to run for governor of California solely to be exempted from having to post truthful ads (or, more realistically, solely to make a protest-point about what he thinks about Facebook's political ads policy). Facebook has already said that they won't allow him to run false political ads on its platform, and Hamptom says he's "considering legal action." Any such legal action would flop, thanks to CDA 230. Once again, content moderation at scale runs into lots of challenges and obstacles, no matter what you do -- and it's particularly fraught in the political advertising context.

Facebook execs have tried to make this point recently, though it's doubtful that anyone is truly convinced:

Anyone who thinks Facebook should decide which claims by politicians are acceptable might ask themselves this question: Why do you want us to have so much power?

In our view, the only thing worse than Facebook not making these calls is for Facebook to make these calls.

Part of the issue is that everyone is conflating a few different issues -- including the powerful position these companies have within the advertising ecosystem, the ability of politicians to target ads, the success of those advertising campaigns, and the nature of truth itself. Each of those are challenging issues, and not all solutions work the same for each -- yet they all get lumped together. And, inevitably, that means a dissatisfying result for all.

But there's another option: which is not to play at all.

Amusingly, the very next paragraph in Facebook's attempted defense of its policy is to try to tie itself to other scrutinized platforms:

Our approach is consistent with companies like YouTube and Twitter. And broadcasters are required by federal law not to censor candidate ads.

To which Twitter has replied: "Nuh uh!" and officially announced it won't allow any political ads on its platform at all:

Twitter is planning to ban political ads from its service globally, the company announced Wednesday via a series of tweets from its CEO Jack Dorsey. The ban will go into effect Nov. 22.

Dorsey said the ban will cover ads about specific candidates and issues — the broadest possible ban. Some ads will be allowed to remain, including those encouraging people to vote. According to a Twitter spokesperson, news organizations are currently exempt from its rules on political advertising, and the company will release full details on exemptions next month.

It will be interesting to see how this plays out in practice -- and I can already predict that there will be judgment calls about what is and what is not a political ad in the coming weeks and months, and plenty of criticism will be leveled when groups think the company decides incorrectly (one way or the other). Again, this remains something of a no win situation in which lots of people will be unhappy.

However, this brought to mind a larger point that I thought was worth making. One of the trite arguments that people keep making about these companies and the decisions they make is that they're entirely about what will increase revenue -- and that these companies want to accept whatever ads they can to maximize that revenue at every opportunity. But, of course, revenue is only one side of the equation. How much that revenue "costs" is a big deal as well. And, it seems pretty clear that Twitter (while watching what Facebook was going through) decided that the headache of dealing with this question was likely way too big of a "cost," even if it wasn't directly a monetary cost.

For what it's worth, Mark Zuckerberg himself has argued that the revenue from political ads is negligible in the grand scheme of things, so the company could ban them without a significant hit. It has just chosen not to, for whatever reasons (Facebook tries to suggest lofty ideals about "giving people a voice" which seems like utter nonsense, because paid advertising has nothing to do with "giving people a voice.")

I raise all of this to go back to my recent paper on "Protocols, Not Platforms." One of the most common criticisms I've heard of that paper is that none of the big social media companies would ever adopt such a system, because it would likely mean giving up control and some amount of revenue (advertising or otherwise). In the paper I try to argue that this is not necessarily true. For one, there are some possible new business models that could replace advertising, but more importantly the costs of continuing to deal with complaints about content moderation are likely to continue to grow at an increasing rate -- and some of those platforms may decide that it's just not worth it any more. And then they may decide that they need to pick another path -- and moving to a protocols-based solution, in which they shift the power away from their own centralized control, and out to the ends of the network, could become much more appealing.

And that's why I find Twitter's decision here quite interesting. It's not going nearly as far as I hope these companies will go eventually -- but it does show that the headaches created by setting themselves up as arbiters (or not) of truth might be so painful and costly that companies will look for ways to get out of the business altogether. That's actually an encouraging sign. It's a hell of a lot better than a company insisting that it can somehow magically deal with this mess and choose which political ads are okay and which are not without pissing everyone off.

Filed Under: advertisement, benefit, content moderation, content moderation at scale, cost, political ads, revenue, social media, truth
Companies: facebook, twitter