POLITICO’s weekly transatlantic tech newsletter uncovers the digital relationship between critical power-centers through exclusive insights and breaking news for global technology elites and political influencers.

By MARK SCOTT

Send tips here | Subscribe for free | View in your browser

SALUT DE PARIS.This is Digital Bridge, and I’m Mark Scott, POLITICO’s chief technology correspondent. This week’s newsletter comes from the City of Light, where I’m on a panel this afternoon (4:30 p.m. CET / 10:30 a.m. ET) on the global race to create AI rules. You can watch along here.

Logistics note: I’m away for the next two weeks, and my POLITICO colleagues will be quarterbacking Digital Bridge. Be gentle with them.

Ready? Let’s do this:

— If we want social media to be more accountable, we first need to create the right tools to know what’s going on.

— The United Nations wants to wade into the thorny issue of artificial intelligence. We should all be very wary.

— Canada is the latest skirmish between platforms and publishers over online content. It won’t be the last.

A ‘CLEARINGHOUSE’ FOR SOCIAL MEDIA

I’M GOING TO CHANNEL MY INNER GEN-Z INFLUENCER and say that everyone needs a side hustle. Mine — since last September — has been as a visiting fellow at Brown University’s Information Futures Lab, where I’ve focused on how to turn Europe’s new online content rulebook into something that people can actually use. Specifically, my research, based on more than 50 interviews with regulators, academics, public health officials and others around the world, relates to how best to give outsiders better access to social media companies’ data for improved transparency, accountability and, inevitably, better policymaking.

But data access is boring, I hear you say. Yes, it is. But what the last seven years, dating back to the 2016 election cycle, clearly demonstrate is that we still have a very blurry understanding of how social media really works. There are limited ways to see what’s happening on these platforms, at scale. To fix that, the European Union wrote mandatory data access provisions into its Digital Services Act — the first, and currently only, legislation to legally force companies like Facebook, Google and TikTok to open themselves up to outside scrutiny. For me, it’s the first step to improving everyone’s social media feeds. You can’t fix what you don’t know, amirite?

What became quickly apparent via my fellowship, though, was the current data access ecosystem is seriously flawed. Regulators — which have their own enforcement powers to review social media data — don’t know where to start. Academics, who currently have the best access to these platforms, often compete with each other for resources and/or are beholden to ad hoc relationships with the companies. Civil society groups either lack the funding to do this work or the technical expertise to do it well. Journalists mostly scramble around in the dark, whipsawing between one-off projects that don’t really move the needle. Platforms themselves struggle with internal power dynamics that have thwarted engagement with outsiders.

What’s needed, then, is a way to do social media data access work at scale and make it available to as many people as possible. That would democratize who can do this by removing barriers that currently stifle accountability; reduce costs as people don’t have to replicate existing data collection practices already carried out by others; and allow better analysis across multiple social media networks (where the real dangers lie) by providing a one-stop-shop for those in need of data access. It goes without saying this work must uphold people’s privacy rights; protest proprietary corporate information; and avoid capture by commercial/government interests seeking to track people online.

So that’s the problem — one that Europe’s new content rulebook (albeit flawed) is uniquely placed to solve. That goes for both those within the 27-country bloc and those elsewhere, given the Digital Services Act’s data access provisions can apply to non-EU groups. The answer, for me, is what I’m calling a “social media clearinghouse,” an underlying digital infrastructure that would bring together platforms’ accessible data into a universal database available to vetted researchers, civil society groups and others. It would take care of the technical layer of this work, so organizations can get on with what they do best: providing a level of accountability and transparency to what happens on social media.

I can already hear you shouting: “You want to create a global database to track everyone’s social media activities? Hard pass.” I get that. The clearinghouse would have to run in conjunction with a strict, independent vetting procedure — one already being created via the European Digital Media Observatory (more on that here) — to limit who can access such infrastructure. This isn’t about creating a global surveillance regime. Think about it more like a CrowdTangle 2.0, a replacement for the cross-platform social media analytics tool owned by Meta that’s indispensable for existing social media accountability work, but which is falling apart, mostly out of neglect.

What my Brown fellowship made clear to me is the current work in this area represents less than the sum of its parts. The limited resources available for social media transparency work is balkanized in a million ways. Brussels has, uniquely, laid out the ability to force greater access to platform data with its new content rulebook, potentially for a global network of organizations. But policymakers have done little, if any, work to turn these rules into a reality for anyone beyond their own enforcement investigatory work.

A social media clearinghouse would fill a much-needed gap by providing the underlying digital data infrastructure to jumpstart a wider array of organizations’ work on unpicking the black box that social media still remains. It’s not going to fix the dumpster fire that parts of this online world have become. But it can provide a universal foundation for what policymakers keep telling me is their goal: to make social media a better experience for all.

THE UNITED NATIONS AND AI

TO MISQUOTE RONALD REAGAN, “The 13 most terrifying words in the English language are: ‘I’m from the United Nations, and I’m here to help (on artificial intelligence).'” Yes, that’s a pretty hamfisted metaphor. But I am incredibly skeptical of any U.N. effort to wade into international digital policymaking. Mostly, that’s because such efforts give authoritarian governments like China, Russia and Saudi Arabia a seat at the table in how things like AI, online content rules and cybersecurity are shaped. Call me an elitist, but let’s work it out between democracies first.

Still, Thursday marks the sixth installment of the U.N.’s AI for Good summit in Geneva. Overseen by the International Telecommunications Union, a Swiss-based UN agency, the two-day gathering is aimed at harnessing the emerging tech to help meet the international organization’s sustainable-development goals. It includes panels on everything from the technical challenges around regulating generative AI to using the technology to fight climate change. Given the summit started years before AI became the over-hyped beast it is today, it would be hard to claim the U.N. is merely jumping on the bandwagon of the latest hot policy area.

“I’m hopeful,” Doreen Bogdan-Martin, the International Telecommunication Union’s American head, told me when I asked her how optimistic she was the U.N. could successfully bash international policymakers’ heads together, given the litany of failed global digital initiatives that pot mark the internet superhighway. “We sort of don’t have a choice. We have to try to make this work.” I’m not sure if the “What else can we do?” argument is a strong one. But Bogdan-Martin is right in that global cooperation on all facets of AI is needed, and needed now.

Luckily, the U.N. isn’t starting from scratch. Back in 2021, UNESCO — another U.N. agency — published its AI Principles, underlying ethics guidelines around things like upholding people’s right to privacy; the proportionate use of AI to carry out specific tasks; and a need for accountability and transparency baked into how these opaque systems are developed. All of UNESCO’s 193 members (but, pointedly, not the United States, which is about to rejoin the group) backed the proposals, including countries like China.

And that’s where things start to fall down. I have my personal views on allowing Beijing to participate in such global digital policymaking fora. But if countries agree to certain underlying AI ethical principles, how do you square that with China’s aggressive roll-out of the technology, especially around its controversial social credit system?

“Well, China signed (the principles),” Gabriela Ramos, UNESCO’s assistant director general for the social and human sciences who helped to negotiate the agency’s recommendations, said when I asked her if China would abide by UNESCO’s AI Principles. “We take it at face value that if countries sign up to the recommendations, we are expecting that they will implement them.”

Again, not exactly the strongest of statements. But Bogdan-Martin, the ITU boss, made it clear the era of self-regulation (looking at you, White House, and your ongoing meetings with industry about such efforts) is over. “Business, alone, can’t be self-regulating,” she said. “There’s a need to have governments engaged. There’s a really important role for the U.N., for academia, and for civil society.” I agree with her that governments need to roll up their sleeves and set some ground rules. I’m just not so sure the U.N. — with the inevitable complex geopolitics that comes with it — is the right place for those discussions to take place.

BY THE NUMBERS

PLATFORMS VS. PUBLISHERS, CANADA EDITION

TO OVER-GENERALIZE OUR COUSINS FROM THE NORTH, Canadians are typically pretty chilled-out people. So when Justin Trudeau and his officials attack the likes of Meta and Alphabet for threatening to take down news content from their platforms, you know something has gone wrong. This is all in response to local legislation, known as the Online News Act, that will force the tech giants to pay Canadian publishers when their content appears on their sites. The Canadian government estimates it may bolster national media outlets’ coffers, annually, by $250 million by mandating Facebook and Google to negotiate such commercial deals.

This follows separate (successful) efforts in Australiawhere local publishers are now pocketing an estimated $150 a year in similar platform payments. In Europe, national governments are now strong-arming tech companies to similarly hand over blockbuster fees via the bloc’s copyright directive, which allows outlets to charge whenever their content appears on these platforms. (Disclaimer: Axel Springer, POLITICO’s German owner, was a vocal supporter of those provisions.) In California, state lawmakers are considering a similar Canada/Australia-style model.

There’s a lot to be said about the lobbying efforts from cash-poor publishers seeking to tap wealthy platforms for additional revenues when their underlying advertising business is crumbling. But it’s also true tech giants gain massively from media outlets’ content that appears on their sites. Is that dynamic unsustainable? Probably. But it’s worth remembering both sides are hard-nosed lobbyists and the answer, inevitably, is somewhere between publishers charging platforms for their content and tech giants using this material for free.

WONK OF THE WEEK

NOW THE ENFORCEMENT POWERS OF CALIFORNIA’S privacy rules are in place (as of July 1), it’s time to focus on Michael Macko, who was just appointed the first deputy director of enforcement at the California Privacy Protection Agency.

The University of Pennsylvania law school graduate has a mix of public and private sector experience. Most recently, he was a senior lawyer at Amazon, and previously was both an assistant U.S. attorney at the U.S. Department of Justice and a trial attorney at the U.S. Securities and Exchange Commission.

Overnight, Macko will become one of the most powerful privacy enforcers in the U.S. (albeit his work will have to be done in conjunction with the state’s attorney general.) That includes a likely focus on how mobile apps are complying with the Golden State’s de facto national privacy standards, as well as how companies’ use of AI aligns with the new rules.

THEY SAID WHAT, NOW?

“The United States has fulfilled its commitments for implementing the EU-U.S. Data Privacy Framework announced by President Joe Biden and European Commission President Ursula von der Leyen in March 2022,” said Gina Raimondo, the U.S. commerce secretary, after the U.S. Department of Justice approved the surveillance practices of EU member countries and American intelligence agencies updated their own guidelines to comply with a 2020 decision from Europe’s highest court that U.S. data protection safeguards were not sufficient to uphold the bloc’s fundamental privacy rights.

Those steps mark the final stages required by Washington before Brussels approves a transatlantic data-transfer deal, which may come as early as Tuesday or Wednesday next week.

WHAT I’M READING

— The U.S. federal government is already using artificial intelligence in myriad ways. Luckily, the National Artificial Intelligence Initiative Office has brought them all into one place. Take a look here.

— The United Kingdom has its own regime to regulate “online safety.” The country’s regulator in charge of these efforts has outlined exactly what its plans are, and over what time period.

— Germany’s federal cartel office won its case against Meta over claims that it abused its dominant position to unfairly favor its own services via the collection of people’s online data. Read the decision here.

— A U.S. judge ordered vast parts of the federal government not to meet or coordinate with social media companies over allegations officials and tech executives were censoring or suppressing people’s protected speech online. Read the ruling here.

— Meta released Threads, a Twitter alternative, that had already signed up millions of users within its first hours. Unfortunately, it appears you can’t delete your Threads account without also removing your Instagram account. Awkward.

— China is moving ahead with its own efforts to regulate generative AI. Yirong Sun and Jingxian Zeng unpick the draft proposals for the Future of Privacy Forum.

SUBSCRIBE to the POLITICO newsletter family: Brussels Playbook | London Playbook | London Playbook PM | Playbook Paris | POLITICO Confidential | Sunday Crunch | EU Influence | London Influence | Digital Bridge | China Watcher | Berlin Bulletin | D.C. Playbook | D.C. Influence | Global Insider | All our POLITICO Pro policy morning newsletters