Authoritarian Tech

It has been 15 days since the launch of Threads and the media razzle dazzle about it has become impossible to avoid. Meta’s new Twitter-like service, which various outlets have casually termed  a “Twitter killer,” racked up more than 30 million users by the end of its July 5 launch date, and topped 100 million by July 10. Whew! Zuck must be getting champagne flute emojis in record numbers.

But hang on – let’s pause for a quick calendar check. Threads debuted on July 5. Sure, it was just after a holiday weekend in the U.S. But it also came on the heels of something much more consequential for the company. On July 4, the European Court of Justice issued a historic ruling that undermines Meta’s legal justification for serving targeted ads in the EU and empowers competition regulators to go after the company on data protection grounds. Legalese aside, this ruling could effectively blow up Meta’s business model in the region, and has put the company’s operations there into serious jeopardy. I recommend Jason Kint’s breakdown if you want the legal nitty gritty. But the bottom line is that Meta is in big trouble for collecting tons of user data without sufficient consent and then mixing and matching it across its services. This is the engine driving the creepily precise ad targeting jiu-jitsu for which Meta is so well known and upon which its business model is based. 

If Zuck is emoji-toasting Threads’ success with one hand, chances are good that with his other hand he is feverishly texting the heaviest of his legal and political heavyweight pals like never before.

The effects of the ruling are already playing out. Norway’s data protection authority stepped up last Friday and temporarily banned Meta from engaging in “highly opaque and intrusive monitoring and profiling operations” (AKA serving targeted ads). Beginning on August 4, Facebook and Instagram will be allowed to serve ads to users, but those ads may only draw on information that appears in the “about” section of users’ public profiles.

If Meta fails to comply with Norway’s ban, it will be stuck with a daily fine just shy of $100,000. While that alone won’t make much of a dent in Meta’s bottom line, it’s part of a bigger picture that almost certainly will. Other EU states are bound to follow suit. Plus there’s the 390 million euro fine that Meta is already facing in Ireland, stemming from the strategic litigation efforts of Austrian privacy lawyer Max Schrems. Mix this all together with the obvious anxiety that it will cause among investors, and you have yourself a serious problem.

If you think Zuckerberg isn’t nervous about it, look at Threads just one more time. Unless you’re in Europe, in which case, you can’t, because Threads hasn’t launched there yet. Instagram head Adam Mosseri says this has to do with the company’s efforts to comply with the EU’s Digital Services Act. But the timing is conspicuous. They may actually be running scared.

Researchers in Switzerland recently rolled out an AI model that they say can detect homosexuality by scanning a person’s brain. AI ethics and LGBTQ rights experts say the model is dangerous. Coda’s Isobel Cockerell has the story.

GLOBAL NEWS

The launch of Threads is also a reminder of the risks that Meta continues to take in monetizing public conversations. Computer and social science researchers have shown again and again that the data collection practices that lie at the heart of what I will now go ahead and call Meta’s surveillance advertising business model are a big part of what makes this a dangerous proposition. The model’s success depends on attention – views, clicks, likes, shares and comments – and any quasi-public space that does this is bound to run into trouble when it comes to political speech, elections, conflict zones, and any other situation where harmful speech or disinformation can go viral and really hurt people. Hate speech in a vacuum is hateful and might hurt the handful of people who see it. But hate speech amplified by an algorithm can reach and inspire millions. We need only look at Myanmar, Ethiopia, or most recently Sudan to understand that it can be deadly.

Meta seems hopeful that we’ll somehow forget this and embrace Threads as a “positive and creative space to express your ideas.” Mosseri has actually said he hopes to keep “politics and hard news” off of Threads. In an intro post, he wrote: “my take is, from a platform’s perspective, any incremental engagement or revenue they might drive is not at all worth the scrutiny, negativity (let’s be honest), or integrity risks that come along with them.” It must be a real drag having to answer to the public and investors about the fact that Meta’s platforms regularly amplify threats of violence and hate speech that has led to real-life harm. The New Republic’s Molly Taft mock paraphrased Mosseri with this zinger of a tweet: “head of instagram says: pls don’t be a downer on threads, we don’t want to have to spend money on fact checking, also no one bring up Myanmar thx.” Buuurn.

And Meta is feeling the heat elsewhere too. Uganda’s parliament slapped a 5% levy on Meta alongside several other Big Techs last week, as part of a new tax law that will apply to “non-resident providers of digital services” like Meta, Twitter, Amazon, Netflix and Uber. The law’s proponents say it will boost public revenue and that it’s overdue from foreign companies that turn big profits in the country. State Finance Minister Henry Musasizi put it this way: “We are looking at the money obtained by the supplier of these services. The money for Uber goes to California; the man makes money but doesn’t pay taxes.” Fair enough. 

And Uganda is not alone in this pursuit – Nigeria and Kenya have both introduced similar tax laws targeting digital services in recent years. But can policymakers find a way to keep the tax from being passed on the consumers? Social media and taxation haven’t mixed well in the past. Uganda’s 2018 “social media tax,” that President Museveni (in power then, in power now) defended as a way to curb “gossip” on social media, forced Ugandans to cough up a daily levy before using social media mobile apps. The law sparked widespread protests, and in addition to violating net neutrality principles, it left poor Ugandans facing a 10% increase in the cost of getting online. A year after it passed, internet use among Ugandans had fallen by 30%.

Lithuania is trying to bring Big Tech to heel too, on the issue of political bots. This week Coda reporter Amanda Coakley and I teamed up on a story about proposed amendments that would criminalize “disinformation, war propaganda, [content] inciting war or calling for the violation of the sovereignty of the Republic of Lithuania by force” from “automatically controlled” accounts. Bigger countries have seen some success using a similar approach, but I wonder if Meta and the rest will be willing to comply in Lithuania, home to just 2.8 million people.

WHAT WE’RE READING

• While European governments are doing plenty of talking to U.S. tech giants, U.S. government officials are currently unsure if they can talk to those same giants themselves. This is because a federal judge in Louisiana recently ruled that state attempts to coerce or otherwise influence companies’ actions is a violation of the First Amendment. Ruh roh! I highly recommend this breakdown of the decision by CJR’s Matthew Ingram.

• Could AI be used to automatically interpret and enforce the law? I’m sorry to say that two very smart people who I know and trust think that it could. Bruce Schneier and Jon Penney set the dystopian scene and then back it up with proper intel and research in this zinger for Slate’s Future Tense vertical.