Archive – 15 Sep 2023

Canada Enacts Forced Media Payments Law Setting Up Platform Showdown

Report from the CBC

In Brief – Canada’s Bill C-18, the Online News Act that will force Google and Meta to pay Canadian news outlets when their media content is posted on the digital platforms, has been enacted into law and will go into effect in six months. The Department of Canadian Heritage will draft regulations specifying the application of the act and provide implementation guidance. The bill is modeled after legislation enacted in Australia in 2021. The platform giants are directed to negotiate payment deals with the media companies, and absent agreements, the parties face off in a binding arbitration process led by a Canadian regulator. In Australia, and since, Meta and Google have tried to draw lines in the sand on media payments legislation. Meta rejects paying for media content posted by users, including by the news media companies themselves. Google has rejected paying for simple internet search links. During recent months, both Meta and Google have admitted to testing changes to their Canadian services to cut off the use of Canadian media content. Meta seems most committed to making a drastic change, releasing a statement following final approval of the bill saying, “We have repeatedly shared that in order to comply with Bill C-18 … content from news outlets, including news publishers and broadcasters, will no longer be available to people accessing our platforms in Canada.” The company indicated that changes would be phased in rather than be implemented immediately.

Context – In recent years, Meta and Google created curated news services on their platforms to pay hundreds of millions of dollars to media companies. The goal was to reduce political heat. Google is still pursuing that strategy. Recent expansions included announcing that the Google News Showcase will open in the US. Meta appears to be going the other way. They claim most social media users prefer TikTok-style entertainment and their focus will shift there. “News” also brings misinformation, angst, and strife. Meta turning off news in Canada will be big news. And a similar bill has momentum in California.

Google Faces In-Car Services Complaint from the German Competition Authority

Report from Bloomberg

In Brief – The Bundeskartellamt, Germany’s national competition authority, has announced that it will investigate Google policies related to its automotive infotainment services platform that it believes are anticompetitive. Google Automotive Services offers carmakers a bundled set of capabilities including Google Maps, the Google Play app store, and the Google voice assistant that allows drivers to have voice-based control of their vehicle infotainment system as well as messaging and telephoning. The regulator objects to the bundling and also raises concerns with Google practices to encourage car manufacturers to preference the Google Assistant as the only voice assistant, or default voice assistant, on infotainment platforms. The investigation is being undertaken under the regulator’s new legal authority to proactively address competition threats from digital companies “of paramount significance on competition across markets” in any market they occupy. Google is also the subject of a Bundeskartellamt investigation into the terms and conditions it imposes on its Maps platform.

Context – The Italian competition authority has already fined Google in 2021 for anticompetitive practices on their auto services platform. However, Germany’s law allowing its competition authority to proactively regulate digital giants rather than just pursue antitrust cases proved a harbinger of things that have come to pass in Europe. Amazon, Apple, Google, Meta, and Microsoft have been so designated in Germany. The EU’s Digital Market Act (DMA) is the prime example. It casts a wider corporate net, likely covering a dozen “digital gatekeepers”, and imposes 18 corporate policy “Do’s and Don’ts”. The German regime is more open-ended in some ways, has more than a year head-start, and currently only targets the five. Some German officials have expressed concerns that the DMA is both too limited in terms of the company conduct that is covered, and that the European Commission may not have adequate resources to effectively regulate so many large firms.

Learning that ChatGPT Is Not a Search Engine – Lawyer Sanctioned for Legal Hallucinations

Report from the New York Times

In Brief – A lawyer who used ChatGPT to prepare a court filing that was filled with fabricated cases and precedents, and his partner, were sanctioned in US Federal District Court. At his hearing, Steven Schwartz said he learned about the AI chatbot from his college-aged children and from media articles and told Judge Kevin Castel that he assumed the tool “was, like, a super search engine,” adding that “I did not comprehend that ChatGPT could fabricate cases.” Schwartz, who appeared contrite, admitted that he did not check the chatbot’s work and said, “I continued to be duped by ChatGPT. It’s embarrassing.” Schwartz’s partner was also sanctioned by the court because his name also appeared on the brief.

Context – ChatGPT has shaken up discussions of AI regulations everywhere. It is important to get grounded in the fact that AI is not one kind of technology. Digital surveillance, self-driving vehicles, automated weapons systems, the list goes on and on. All different. In this case, the matter involves generative AI, large language models, and neural networks. Chatbots. A key learning regarding these tools is that super great chatbots are nothing like traditional databases or internet search engines. They don’t store and return fixed data. Instead, they are built to produce realistic, human-sounding responses each time based on chat-style queries. They do this by determining which fragments of text best follow other sequences, all based on a statistical model that has ingested and processed billions and billions of examples pulled from all over the internet. It is entirely possible that they deliver factual accuracy, credible ideas, and responses that seem sensible, even creative. But being “correct” is not the goal of the tools. If you are interested in the technology in the context of “intelligence”, read the summary on page 93 of this report by Microsoft experts, the primary corporate partner of ChatGPT’s OpenAI. They describe how far away the model is from human-style intelligence due to its inherent limitations. It sheds light on why chatbot “hallucinations”, or realistic-sounding fabrications such as those that plagued the brief in question, are basically a design feature rather than mistakes.

US Pressing for Narrow Council of Europe AI Treaty with Private Sector Opt-In

Report from Euractiv

In Brief – The Council of Europe, an international organization focused on human rights and made up of 46-member countries, has been engaged since 2019 in talks to craft the first legally binding international treaty on artificial intelligence. However, with the explosion of interest in AI that accompanied the public rollout of OpenAI’s ChatGPT, the talks have gained importance as a potential forum for some manner of international regulation of the technology. European negotiators reportedly believed that the Council of Europe treaty could be structured to parallel the AI regulatory scheme being developed in the EU through the AI Act. That EU legislation, which will soon be the subject of negotiations between the three EU legislative bodies, is expected to establish comprehensive rules for AI applications. However, the US delegation is pushing to structure the draft agreement in a way that only binds signers to establish national laws governing how the governments themselves use AI, while leaving each country with discretion to extend AI regulations to companies and other private sector users if they choose. The US proposal to set the treaty default as only obliging governments to establish rules for themselves is reportedly backed by the UK, Canada, and Israel.

Context – Is the comprehensive regulatory scheme of the EU’s AI Act on its way to becoming a “global standard”? The news from the Council of Europe’s AI Treaty talks adds to the weight of evidence that despite the media chatter, it’s unlikely anytime soon. ChatGPT bursting on the scene pushed the EU Parliament to propose tighter regulation of general AI “foundational models”, moving away from the legislation’s focus on risky applications themselves. If included in the final legislation, this moves the EU farther from the many governments calling for high-level AI principles not direct regulation. Looking for governments to regulate how they use AI themselves may prove challenging enough. The Transparent Automated Governance (TAG) Act in the US Senate, proposing rules for federal government use of AI, is a good test case.

Judge Weighing Injunction to Stop Feds Pressing for Online Content Moderation

Report from the Wall Street Journal

In Brief – US District Court Judge Terry Doughty is preparing to rule on a motion from the Attorneys General of Missouri and Louisiana requesting a preliminary injunction blocking federal officials from pressing social media platforms to censor legal communications by users. The request is part of a federal lawsuit filed by the two AGs in May 2021 accusing senior Biden Administration officials of pressuring Meta, YouTube, Twitter, and others to suppress content on COVID, vaccines, election integrity, and other issues. The proceedings have resulted in the disclosure of many instances where named officials pressed company executives to more aggressively moderate content. The case tests the limits of the principle that government agencies, bound by 1st Amendment limits prohibiting censorship, cannot coerce private parties, who can restrict speech on their platforms based on their own 1st Amendment rights, to engage in activity the government cannot do themselves. In March, Judge Doughty dispensed with Department of Justice motions to dismiss the suit, raising prospects for a groundbreaking ruling.

Context – There are two threads of the Republican social media “censorship” arguments. One is that the largest platforms are either public venues limited by the 1st Amendment directly, or a type of “common carrier” prohibited by law from policing messages. Laws passed in Texas and Florida, and a lawsuit by Ohio’s AG, pursue this line. The Texas and Florida laws may be in front of the Supreme Court next year. The other thread alleges that the federal government illegally circumvents the 1st Amendment by pressuring the platforms. The MO-LA lawsuit is making that case, as are Republicans on the House Judiciary Committee. While allegations from the period when Donald Trump was President make little sense, Biden officials pressuring platforms to police “misinformation” in 2021 is a more credible delegated-censorship allegation. Outside the US, without the 1st Amendment, government efforts to direct platform content moderation are growing irresistibly in countries as diverse as the EU, UK, Australia, India, and Vietnam.

Some AI Giants Discussing Paying Large Media Companies to Train AI Systems

Report from the Financial Times

In Brief – Some of the largest AI technology companies, including OpenAI, Google, and Microsoft (the major corporate funder of OpenAI) have been meeting with news media executives to discuss copyright issues involved in the training of neural networks that support AI chatbots and other generative AI services. Publishers involved have included News Corp, Axel Springer, The New York Times and The Guardian. The discussions, which are reportedly in the early stages, include ideas such as media organizations being paid a subscription-style fee for their content being used in AI training.

Context – “Neural networks” often “train” using unimaginably large bodies of data, such as most or all the texts, images, or music on the internet. Of course, if one gathers and processes everything available on the internet, it includes immense volumes of copyrighted works. But neural networks do not operate like traditional databases or search engines. They don’t gather, classify, store, or retrieve training data. Instead, they generate new output in response to queries. Of course, everything created may be like something that already existed and was reviewed. A recent hearing of the House IP Subcommittee discussed issues related to copyrighted works in AI training databases, including the merits of a new payment regime like a compulsory license. A key insight was provided by AI researcher Chris Collison-Burch (see the exchange at 1:13:30). He noted that the volumes of content used to train the largest neural networks, literally every piece of available text on the internet, means that payments to any individual copyright holder would be very, very small, because no copyright holder accounts for a meaningful share of all the world’s text. On the other hand, if copyright holders are all paid a decent sum, it would quickly become prohibitively expensive to train systems, at least for small enterprises. The European Parliament’s version of the AI Act includes a first step towards regulating the practice with a transparency requirement for AI developers to identify copyrighted materials used in training.

FTC Sues Amazon Alleging Deceptive Dark Patterns on Prime Subscription Processes

Report from the Wall Street Journal

In Brief – The Federal Trade Commission has sued Amazon in US District Court in Washington State alleging that the e-commerce giant engages in deceptive practices to trick millions of consumers into enrolling in its Prime membership program and dissuade them from following-through on cancelling Prime subscriptions when they desire to do so. Prime costs $139 per year in the United States and there are over 148 million US subscribers. The FTC has engaged in policy discussions for several years regarding “manipulative, coercive, or deceptive tactics” categorized by the term “Dark Patterns” and began its inquiry into Amazon in 2021. The FTC’s complaint reveals that the Prime cancellation process was labeled in internal company documents as “the Iliad flow” in reference the arduous journey in classical Greek literature, and that despite making some changes in April, still required five or six steps. An Amazon spokesperson said that the company makes it “clear and simple” to sign up for and cancel Prime and that “We look forward to proving our case in court.”

Context – The FTC lawsuit comes nearly a year after Amazon agreed to make drastic changes in its European business in the face of similar Dark Patterns charges. In early 2021, a collection of consumer groups filed complaints accusing the company of breaching the EU’s Unfair Commercial Practices Directive and various Member State laws by using complicated navigation menus, skewed wording, confusing choices, and repeated nudging to thwart consumers intentions. The Norwegian Consumer Council portrayed Amazon’s practices in this short cartoon video. The FTC investigation began around the same time. Amazon settled the complaints in Europe last July by agreeing to dramatically simplify the process to cancel a Prime subscription, instituting a simple two-click process in Europe. The EU’s Digital Services, then being legislated, was enacted in late 2022 with a high-level prohibition on dark patterns that focuses on three practices – fake countdown timers, the use of design features to “nudge” users, and hiding information by making it less visible on a website page.

Australia’s Safety Commissioner Threatens Twitter with Hate Speech Fines

Report from the Washington Post

In Brief – Julie Inman, Australia’s eSafety Commissioner, has notified Twitter that it is facing fines of up to Aus $700,000 a day unless it effectively addresses hate speech on the platform. Inman Grant, who worked as Twitter’s top government affairs official in Australia from 2014 to 2016, says that a third of the complaints that the regulator receives regarding online hate on social media are for content on Twitter, and that the complaints have increased since Elon Musk took over the platforms. She questions whether the uptick can be attributed to the company paring back its content moderation efforts, including changing some policies and reinstating some formally banned users. In her announcement, Inman Grant noted that complaints about Twitter were growing globally, including from groups addressing online racism, antisemitism, and hate speech targeting LGBTQ+ communities. Australia’s 2021 Online Safety Act mandates that social media providers take “reasonable steps” to ensure user safety, minimize abuse, and protect users from “abhorrent violent conduct” that can include cyber bullying and image-based abuse. Twitter has 28 days to respond to the notice.

Context – Musk’s defense of “free speech” and criticism of the prior regime’s content moderation practices has upset many progressives and delighted many conservatives. But more relevant to content moderation policy are likely to be his repeated reference to Twitter following “local laws”, which includes the First Amendment in the United States, but means something entirely different in markets around the world, whether India and Turkey where governments criticized as authoritarian ask for critical posts to be taken down, or the EU and Australia. Musk has repeatedly said that Twitter would abide by the legal requirements of the high-profile Digital Services Act, the EU’s new uber-content moderation law, but regulatory challenges to the company’s implementation practices are still to come. National versions in Germany, and now Australia, offer a foretaste, as does the European Commission’s DSA “stress test” being run before the law even comes fully online.

UK CMA Clears Amazon’s iRobot Acquisition

Report from Bloomberg

In Brief – The UK Competition and Markets Authority has cleared Amazon’s proposed $1.7 billion acquisition of iRobot, the manufacturer of the Roomba robotic vacuum devices, announcing that they do not believe that the deal would lead to competition concerns in the UK. Reasons cited in the announcement included that iRobot’s market position in the supply of robot vacuum cleaners in the UK is modest and that it already faces several significant rivals, that Amazon did not have an incentive to undermine rivals on its ecommerce marketplace platform due to the significant losses of fees and commissions it would suffer, and that robot vacuum cleaners are not considered an important input to the emerging “smart home” market. Critics have raised novel data privacy concerns, claiming that Amazon could use the machines’ ability to collect sensitive video and digital mapping data, charges that the company denied and did not appear to be a major component of the CMA review.

Context – Welcome to another round of Antitrust Enforcement Roulette with the US, EU, and UK regulators. Roomba’s market share slide, iRobot’s lack of profitability, and the fact that Amazon did not have its own robot vacuum, all raised questions about the strength of a legal challenge. Yes, there is general opposition to Big Tech acquisitions, but the non-traditional privacy charges regarding in-house Roomba videos and maps have stood out. Similar privacy accusations recently surfaced about Tesla in-car videos. The European Commission is reportedly digging into the issue. Meanwhile, the CMA has appeared particularly aggressive on tech acquisitions, forcing Meta to sell startup Giphy (at a huge loss) and also moving to block Microsoft-Activision, belying to some the UK Government pitch that the country will be a hub of tech investment and innovation. The European Commission is scheduled to announce their initial iRobot decision by July 6. Maybe they will pick up the Bad Cop mantle. The FTC’s Big Tech acquisition focus is likely on the expedited federal court review of its effort to hold off the massive Microsoft-Activision deal.

Tech Antitrust Skeptic the New Top Democrat on House Antitrust Subcommittee

Report from CNBC

In Brief – Representative Louis Correa, a fourth-term Democratic Member of Congress from California, will be the new top Democrat on the House Judiciary Committee’s Antitrust Subcommittee. Correa was a rare dissenting Democratic voice last Congress when the House Judiciary Committee reported out a landmark set of antitrust bills targeting Big Tech practices. The bills, which had bipartisan backing, also faced opposition on both sides of the aisle. Along with relentless lobbying from the digital giants and their business community advocates, opposition from key California Democrats in the House was often cited as a reason the legislation stalled in that body. The shakeup at the top of the Democratic side of the Antitrust Subcommittee was initiated when progressive stalwart Rep. David Cicillini, the lead champion of the Big Tech antitrust campaign, unexpectedly announced his retirement from Congress to become the CEO of the Rhode Island Foundation. Correa has hewed a more business-friendly position on tech regulation and was endorsed by the US Chamber of Commerce in 2022.

Context – Last Congress’s antitrust campaign was seen by many as a real threat to the tech giants because it brought together ascendant progressive Democrats who backed major antitrust reforms with conservative Republicans angered by digital giants. Cicilline’s strategic partnership with conservative Rep. Ken Buck (R-CO) at the top of the Antitrust Subcommittee was the clearest manifestation of that coalition. Nevertheless, growing Republican concerns with regulatory overreach, especially at the FTC, combined with Democratic naysayers, held the bills back. When Republicans won a narrow House majority, Cicilline lost his chairmanship. Then Republican leaders bypassed Buck for the subcommittee helm, installing an antitrust skeptic and reinforcing the impression that Big Tech antitrust reform was off the agenda. Elevating a tech antitrust skeptic into the top Democratic seat completes the House subcommittee leadership reversal. In the Senate, Big Tech antitrust reform backers have reintroduced their bill.

EU Parliament Approves Its Version of the AI Act

Report from the Washington Post

In Brief – The EU continued on its way to regulating Artificial Intelligence technologies as the EU Parliament approved its version of the AI Act. The three EU governing institutions will now try to hammer out a final version by year’s end. The bloc is five years into this effort. The EU Commission and Council of Member States produced similar AI Act versions built around a risk-based framework regulating applications rather than general technology. Stringent rules apply to “high risk” uses that threaten health, safety, and fundamental rights. However, the Parliament’s work coincided with the emergence of ChatGPT, causing a major rethink on regulating “generative AI” and “foundation models” upon which some AI applications are built. Their bill includes stricter obligations for foundational models, requiring generative AI tools to disclose the use of copyrighted material in their training data, an expanded definition of high-risk applications, and stricter treatment of biometric surveillance.

Context – The EU will eventually be the first major market to have a comprehensive AI regulatory regime. And many analysts are saying that the EU AI Act will become a “global standard.” Don’t plan on that anytime soon. First, don’t overlook how the EU Parliament’s changes to tighten regulation of general foundational models rather than the long-developed focus on risky applications moves away from the broader global consensus. The real trend is to try to agree on high-level principles. This is true in the US, at the G-7 (with Japan leading the way), South Korea, and the UK. Even the EU Commission is trying to create a voluntary “AI Pact” of high-level principles. The lack of agreement on what “AI” even means was reinforced when the US-EU Trade and Technology Council created a working group just on AI terms and definitions. Rules for how governments themselves use AI is challenging. Watch the EU bodies fight over government surveillance in the final AI Act. And while the US Congress is definitely not doing broad AI legislation, the bipartisan Transparent Automated Governance Act in the Senate, aiming to set rules for the federal government’s use of AI, is a test case on whether serious, limited legislation is possible.

Amazon Criticized for Dominating Small Online Retailers in Europe

Report from the Retail Insight Network

In Brief – The Centre for Research on Multinational Corporations (SOMO), a Netherlands-based public interest research organization, has released a report alleging that Amazon’s dominance of online commerce marketplaces and ecommerce logistics in Europe allows the company to engage in exploitative treatment of sellers on its platform. The group argues that Amazon’s marketplace is an essential platform for small business online sellers, and that the ecommerce and logistics giant has aggressively ramped up fees charged to sellers at rates exceeding the rates that sellers have increased sales. SOMO claims that Amazon earned €23.5 billion in combined listing fees and logistics services fees from third-party sellers in Europe in 2022, more than triple the €7.6 billion in 2017, and that Amazon has used on-platform “advertising” fees, which allow third-party sellers to pay supplemental fees in order to appear higher in search results, to earn an additional €2.75 billion from EU sellers in 2021. SOMO calls on European authorities to either regulate Amazon as a public utility or break up its different businesses, in particular separating its logistics service from its marketplace.

Context – SOMO’s report on Amazon’s dominance in Europe leading to extractive policies imposed on small business sellers parallels similar work by the Institute for Local Self-Reliance (ISLR) in the US. ISLR argues that while Amazon’s traditional first party retail is a low-margin business, when small businesses sell on Amazon using FBA, it is a very profitable, high-fee, high-margin business for Amazon. This profitability, with total fees on top sellers averaging 50% of sales, is why Amazon uses all its algorithmic levers to push sellers using FBA into the “Buy Box”, a practice that is at the heart of the recently settled antitrust case in the EU, is under investigation in the UK, and could be a top issue if the FTC gets around to challenging Amazon. The company has said that it hopes its settlement with the European Commission, which will include presenting a second Buy Box with alternative delivery, will be a model for its compliance with the newly enacted Digital Markets Act.

European Commission Continues Review of Booking.com – Etraveli Deal

Report from Reuters

In Brief – The European Commission competition authority, which is reviewing Booking.com’s plan to acquire “Online Travel Agency” (OTA) Etraveli for $1.83 billion, has announced its preliminary conclusion that Booking is the dominant hotel OTA in Europe, and that the acquisition may strengthen its position of dominance and reduce competition in a range of travel services markets. Booking announced in 2021 that it planned to acquire the Sweden-based OTA, which was already powering Booking’s then nascent flights’ reservation offering, an important expansion of Booking’s accommodations-focused OTA platform. The Commission is concerned that Booking expanding its ecosystem of travel services through acquisition will increase its bargaining position towards hotels, divert demand from cheaper alternative sales channels, and make it harder for competing OTAs to build and grow hotel OTA businesses. The European Commission is scheduled to decide on the deal by August 30.

Context – Competition officials in the US, EU, and UK have all expressed deep concerns with Big Tech acquisitions, but there is no prevailing theory behind a change in merger reviews, nor a clear trend in decisions. The UK CMA and US FTC took aggressive action to stop Meta deals for small startups. The CMA unwound Meta’s deal for Giphy, while the FTC lost in federal court trying to block Meta’s bid for VR app developer Within. On the far other end of the size spectrum, the CMA is trying to block Microsoft’s mega deal for Activision-Blizzard, with the FTC suing to block it as well. But the EU approved that deal. This Booking bid is interesting because the buyer far smaller than the biggest digital giants but is still likely to be a “gatekeeper” under the new EU Digital Markets Act (and has been designated a VLOP under the DSA as well) and Etraveli is valued at nearly $2 billion, which for comparison’s sake is more than Amazon target iRobot. Neither of the EU’s two big digital bills change the standards to judge whether acquisitions are anticompetitive, so the decision of the Commission will be a sign of how much they don’t want any “gatekeeper” to be able to buy their way to a stronger market position.

Both Epic Games and Apple Appeal the Latest Federal Court Decision. Again.

Report from Reuters

In Brief – As fully expected, both Epic Games and Apple have appealed aspects of the April federal appeals court decision in their epic app-store litigation battle. A three-judge panel of the Federal Ninth Circuit Court of Appeals affirmed the 2021 ruling by Federal District Court Judge Yvonne Gonzalez Rogers that Apple is not an illegal monopoly and did not violate federal antitrust law, but that the iPhone giant did violate California Unfair Competition Law and was ordered to end their anti-steering policies. Like in the trial court, the appeals court decision was largely in Apple’s favor, but not entirely. And once again both Epic Games and Apple have appealed. Apple’s appeal focuses on the imposition of a nationwide injunction over policies regarding consumer payments policies that Apple says are procompetitive and do not violate US federal law, while Epic’s appeal argues its claims implicate the “core purpose” of US antitrust law to foster competition and that the appeals court did not conduct a “rigorous” balancing between asserted consumer benefits and anticompetitive effects of Apple’s practices.

Context – Epic Games and the campaign to regulate the Apple and Google mobile ecosystems put together an impressive coalition for the initial Epic appeal. (Amicus briefs here.) But they basically lost again. The most compelling message from the appeals court was that Apple’s privacy and security justification for their walled garden model is strong. And consumers are aware of it. Finally, Epic did not provide an alternative model that met those needs and compensated Apple for their investments. With US judges proving resistant to political winds, the digital regulation and antitrust regimes in the US and EU are moving very far apart. The EU DMA comes online soon and will likely regulate the big app stores. New in-app payments systems created by Apple and Google will only become more prevalent and important. And it will be interesting to see how Epic’s parallel federal trial challenging Google plays out. While Android has always been more open than Apple’s iOS, it’s less clear how consumers (or a jury) understand it.

Meta to Give Users More Control to Mollify German Competition Regulator

Report from TechCrunch

In Brief – The German Federal Cartel Office (FCO), the country’s competition authority, has announced that Meta has agreed to give users more control over the operations of their accounts, establishing an accounts center that the regulator said will allow customers “to make a largely free and informed decision about whether they want to use Meta’s services separately or in combined form.” For example, users would be given the choice to enable functionalities across popular Meta platforms, such as cross-posting photos on both Facebook and Instagram, knowing that if they choose those options then Meta would use data from both platforms for advertising. On the other hand, if users chose to keep their accounts separate, they would be free to choose that option. The proposal from Meta comes amidst a legal standoff that was kicked off in 2019 when the FCO issued a landmark decision that the social media giant’s data policies of “super-profiling” were an abuse of its market power that the competition regulator could address. The company challenged the regulator’s authority to intervene on a privacy and data matter, winning a stay in German courts and a referral to the EU’s top court. A decision is expected in early July, and Meta’s policy, which will be applied globally rather than just in Germany, appears set for a decision in favor of the regulator.

Context – The German FCO’s effort to regulate Facebook’s data use and advertising practices as a competition abuse was a landmark in the EU pushback against the digital giants. But it has largely been overtaken by events. In 2021, due in part to the years it took to conclude antitrust cases, the FCO was given novel and broad standing regulatory authority over giant digital platforms it determined were “of paramount significance on competition across markets” and has since decided that Amazon, Apple, Google, Meta, and Microsoft met that standard. The EU’s Digital Markets Act takes that model of ongoing regulation rather than case-by-case antitrust enforcement EU-wide, with 18 high-level company “Do’s and Don’ts” for what may be a dozen “digital gatekeepers”.

Directed by Congress, Texas’s Google Antitrust Suit Looks Like It Is Heading Back to Texas

Report from Bloomberg

In Brief – Texas and 16 other states pursuing a federal antitrust suit against Google’s digital advertising business have won the latest round in their multiyear effort to pursue their case in federal court in the Eastern District of Texas. The federal Judicial Panel on Multidistrict Litigation (JPML) ruled that the State Antitrust Enforcement Venue Act of 2022, legislation enacted last December to give State Attorneys General the right to prosecute federal antitrust suits in the district court of their choosing, an authority previously reserved for the US Department of Justice, applies to pending state-led antitrust suits. Therefore, the Panel said it was required to grant the motion for remand filed by Texas. (Note: In a late development, the panel accepted a motion from Google to stay the transfer and rehear arguments at the JPML’s July 27 hearing with full briefings due by July 6.) The Texas-led advertising-focused antitrust suit was filed in late 2020, Google aimed to move it to federal court in California, the AGs argued for Texas, and the JPML chose to consolidate it in the Southern District of New York with several similar private complaints. If the latest JPML decision stands Google will be battling antitrust lawsuits targeting its digital advertising practices in federal courts in New York, Texas, and Virginia.

Context – Google has been the top digital ad business since it acquired display ad platform DoubleClick in 2007. They have since drawn antitrust scrutiny in many markets, including the UK, EU, and Australia. The 2020 Texas-led AG suit was joined by a similar DoJ-led suit (that included six more States) in January. Google’s effort to move the DoJ complaint to New York was quickly turned down, and the pending remand back to Texas illustrates that while most of last-Congress’s Big Tech-antitrust fervor came to naught, giving State AGs the ability to pick their federal court venues will impact the evolution of case law, likely by engineering more circuit splits. On the substance, all these cases will increasingly need to address the fact that the overall digital ad market shares of Google, and number two Meta, are falling, with Google’s now below 30 percent.

The EU Submits an Antitrust Complaint Targeting Google’s Top Moneymaker – Advertising

Report from the New York Times

In Brief – The European Commission competition authority has filed a formal antitrust complaint against Goolge alleging that it has engaged in years of anticompetitive conduct in its massive digital advertising business and indicated that it would likely press for a breakup rather than promises of better behavior in the future. The charges come two years after the EU’s antitrust enforcers began their investigation of Google’s digital advertising business, which accounts for the bulk of the company’s revenues. The Commission argues that the digital “adtech” industry is built around three digital services: publisher ad servers used by publishers to manage their advertising space; ad buying tools used by advertisers to manage their digital ad campaigns; and ad exchanges where publishers and advertisers interact in automated auctions to buy and sell display adds. It says that Google operates in all three markets and is the dominant business in two, publisher ad servers and ad buying tools. They argue that Google has used that dominance “since at least 2014” to preference its ad exchange, which they believe likely foreclosed rival ad exchanges and reinforced its “ability to charge a high fee for its service.”

Context – Google emerged as the preeminent adtech giant following its acquisition of DoubleClick in 2007 and has drawn major antitrust scrutiny in numerous markets. This EU complaint parallels charges from US State AGs led by Texas and the US Department of Justice. Those trials might be as early as next year. Defining the relevant markets, measuring shares, and trying to prove market power, are key to traditional antitrust analysis, and that will again be the case. The conduct behind these complaints all stretches back years, many when Google’s shares of the overall adtech market was higher than today. Now, Google’s share is falling, with reports pinning it below 30% of the overall market. Therefore, antitrust enforcers will focus on narrower adtech service markets, arguing that the company holds much larger shares and exercises great influence. Also, Google arguments that its adtech rates are falling may highlight the key differences between US and EU competition law.

Google Finally Launches Its News Showcase Service in the US

Report from the Wall Street Journal

In Brief – Google has announced its News Showcase service will finally open in the United States. Google’s News Showcase is a licensing program begun in late 2020 to pay publishers for content that appears in a curated news offering and allows users to access some media content normally held behind media company paywalls. The digital giant created the program in a transparent effort to reduce political heat generated by media companies complaining to their government officials that digital ad giants Google (and Facebook) were decimating their business. When it launches in the US this summer, Google says that it will include more than 150 news publications from 39 states, most of which it describes as “local or regional.” Google claims that more than 2,300 publications across 22 countries are now participating.

Context – Newspapers have been complaining for two decades that the internet ruined their business model and lobbying governments worldwide to force the largest digital ad platforms, namely Google and Meta, to pay them more money. Australia and France have had the most success compelling payments, and recent efforts in the US Congress, State of California, and Canada, have generated news. Both digital platforms have drawn policy lines in the sand. Google rejects paying for basic search links. Meta rejects paying for news content posted solely by users, which includes media companies themselves. However, while both companies created curated news services in recent years and paid hundreds of millions of dollars to media companies to reduce pressure, they are now going in opposite directions. Along with its upcoming expansion in the US, Google is also growing media payments in Europe. Meta is going the other way. They claim social media users increasingly don’t want news stories and prefer TikTok-style entertainment. “News” also brings a range of ills. So, Meta responded to recent progress by media company bills in Canada and California by threatening to simply prohibit posting news media stories on Facebook and Instagram in those places. That would be an interesting development.

FTC Asks Federal Court for Injunction to Block Microsoft-Activision Acquisition

Report from the Washington Post

In Brief – The Federal Trade Commission (FTC) has asked a federal court in the Northern District of California to temporarily block Microsoft from completing its $69 billion acquisition of Activision Blizzard until the agency concludes its regulatory challenge of the deal. The acquisition, announced in January 2022, has faced antitrust reviews globally. The FTC, which along with the Biden DoJ, has taken a very skeptical view of acquisitions in general, especially by Big Tech, formally opposed the deal last December. The agency intended for its case to first be adjudicated in its administrative court system, which is scheduled for a trial in August. Although the FTC often petitions a federal court to delay the completion of an acquisition it plans to challenge while it does its work, it did not initially do this with the Microsoft-Activision deal, reportedly believing that reviews in top foreign markets would keep the deal from closing. However, regulators in many major markets have since approved the deal, in particular the EU. The UK CMA has blocked the deal, a decision Microsoft is appealing, and the company is exploring options to close the deal while they pursue that appeal. If a federal court issues a temporary restraining order to stop the deal closing until a decision on a preliminary injunction is reached, it will lead to a federal court trial where Microsoft and Activision can challenge the FTC’s substantive case.

Context – The Microsoft-Activision mega-deal is highlighting the new era of sequential national acquisition reviews, especially with the roulette of the US, EU, and UK, which is calling into question established norms of competition policy. Going after small startup deals like Meta-Giphy, by the UK CMA, or Meta-Within, a federal court loss for the FTC, is one thing. But Microsoft is engaging at the level expected with the biggest Big Tech deal ever. Over 30 competition regulators have signed off globally, with the EU being the most impactful. The FTC is coming off a loss in federal court at this stage on Meta-Within. And Microsoft appears to be testing the waters on how a tech giant reacts if just the UK blocks a deal.

EU Council of Member States Agree on Platform Worker Legislation

Report from Reuters

In Brief – After months of negotiations, the European Council has agreed on a version of the Digital Labor Platforms Directive to regulate Gig-style work platforms in the EU. The European Commission proposed an initial draft in December 2021, and the European Parliament approved their version of the legislation in February. The member states have been divided between governments aiming to strictly regulate Gig-style work and those championing the greater flexibility of independent work models. A top goal has been standardizing when platform workers are considered platform employees with applicable rights and benefits. The Council agreed that workers should be considered employees when they meet three of seven criteria: when the platform determines the maximum remuneration; when it requires certain standards of appearance or behavior; when it supervises the work electronically; when it limits workers’ freedom to determine their hours or absences; when it limits their freedom to reject tasks; when it limits their freedom to use subcontractors; or when it limits their freedom to build their own client base or work others. Rules related to personal data used by labor platforms were also contentious. The Council prohibits labor platforms from using “automated monitoring or decision-making systems” to process personal data relating to the worker’s “emotional or psychological state,” but is less restrictive than the Commission limiting data collection to strictly necessary for the performance of the contract between the platform and the worker. The three bodies will now negotiate a final version.

Context – Unlike in the United States, where labor-backed initiatives to reclassify Gig-style workers have been stymied since late 2020, the EU has been plodding forward toward goals generally supported by labor advocates. However, Left-Right divisions on the value of worker independence, flexibility, and new work models continue to threaten the EU legislative process. And in the US, court challenges from businesses are likely if the Biden Administration Department of Labor moves to enforce new worker classification rules.

Florida Passes Novel Republican Privacy Law Targeting Big Tech

Report from Bloomberg

In Brief – Florida Governor Ron DeSantis (R) has signed a unique hybrid data privacy and Big Tech regulatory bill creating new consumer rights to control how data is handled by the biggest tech companies and provisions aimed at alleged online censorship of conservative views. The data privacy sections, which largely follow laws enacted in other states, include rights to confirm, access, correct, and delate personal data, opt out of data being sold or used in targeted advertising, and restrictions on collecting children’s personal data. One major difference from laws enacted in other states is that the new standards not related to kids data only apply to very large digital companies, including businesses with $1 billion in revenues with half coming from digital advertising, companies that operate smart speaker systems, and those operating the largest app stores. Privacy advocates argued that many data brokers and other digital businesses would be exempt. An even more novel addition were sections related to alleged viewpoint censorship, a national conservative political issue unrelated to privacy, including requiring large search engines to disclose if they prioritize results based on ideological views, and generally prohibiting state and local government employees from coordinating with large social media companies on content moderation.

Context – Florida is the tenth state to enact data privacy legislation since California in 2018. Success is linked to Democrats pulling back from insisting on class action lawsuits being part of the enforcement mix, which saw laws enacted in Democrat-controlled states like Virginia, Colorado, and Connecticut in 2021 and 2022. The latest string of wins has been in Republican-led states. Texas, the other Republican bellwether, looks set to be next. It will be the first state measure to use US Small Business Administration size standards, which often include a 500-employee threshold. On viewpoint censorship, Republican leaders from both states will be in the spotlight defending Florida’s and Texas’s landmark social media content moderation laws in federal courts, potentially to include the US Supreme Court.

Microsoft Actively Pressing Their Activision Acquisition Case in the UK

Report from Reuters

In Brief – Microsoft continues to press the UK Government to accept its acquisition of giant game developer Activision-Blizzard, with company President Brad Smith heading to London for meetings and events, including with finance minister Jeremy Hunt. The Competition and Markets Authority (CMA), the country’s antitrust regulator, blocked the $69 billion takeover in April, rejecting Microsoft’s promises to guarantee access to Activision games on a wide range of platforms and formats. Those offers were key to winning over the EU, Japan, South Korea, and over 30 other competition regulators. Smith was very critical of the CMA decision and said that it sent a global message that the UK was not receptive to technology investment and innovation, which earned him a swift rebuke from top UK Government leaders. At a tech leader event on his recent tour, Smith was more conciliatory, saying he had always been “bullish on the United Kingdom as a great place to live, to learn, to build,” and that “if the UK wants to impose regulatory requirements that go beyond those in the EU, we want to find ways to fulfil them.”

Context – The new era of oft-disjointed sequential acquisition reviews, especially by the US, EU, and UK, is calling into question established norms of competition policy and the rule of law. Multiple regulators review the same deal and sometimes appear to partner and venue shop in ways that are causing heartburn among business advocates. Microsoft-Activision was always going to face reviews in all three markets (and more). The UK CMA setting a mark for aggressiveness on tech deals, despite the UK being the smallest of the three markets (by far), and the EU’s willingness to approve deals with conditions, is calling into question the UK’s pitch to be a tech global business hub. In the US, the FTC and DoJ are aggressively challenging deals, but their record in court is spotty and there is little bipartisan support to change federal merger review law. Instead, questions are being raised about anti-merger US regulators working with overseas regulatory partners to torpedo deals that would more likely prevail in US courts.

France Enacts Comprehensive Regulation of Online Influencer Industry

Report from EuroNews

In Brief – The French national legislature has enacted legislation regulating online “influencers”, defined as “individuals or legal entities who, for a fee, mobilize their notoriety with their audience” to promote goods and services online. There are reported to be upwards of 150,000 paid influencers using social media platforms in France, and they have been criticized from across the political and cultural spectrum for promoting unhealthy lifestyles, scams, and frauds. The new law, which passed with widespread backing, prohibits promotion of a limited list of practices, including cosmetic surgery and ending a doctor prescribed course of treatment, and heavily regulates the promotion of medical devices, nicotine products, and gambling-related services. Promotional images, such as for cosmetics, must disclose whether they have been retouched or use a filter. The measure also regulates influencers’ agents, requires EU-based influencers to hold liability insurance, non-EU influencers to establish a legal representative in the bloc, and platforms to accept and act on notifications of frauds and other deceptions.

Context – The global social media “influencer” industry was valued at $16.4 billion in 2022. In the US, the Federal Trade Commission has been working to provide truth-in-advertising direction to the burgeoning sector since 2017. It is in the process of updating its “Endorsement Guides” for influencers, advertisers that hire them, social media platforms that carry them, and businesses that prepare or present online consumer reviews. The agency is also proposing to move beyond guidelines and issue new regulations regarding unfair and deceptive practices in digital marketing, including regulating fake reviews, paying for positive (or negative) reviews, suppressing negative reviews, or buying or selling followers, subscribers, views, or other means of online influence. The UK Competition and Markets Authority updated its guidance for online influencers last December and the government’s draft Digital Markets, Competition and Consumers Bill includes a section aimed at policing fake online reviews.

Another House Committee Investigating Lina Khan’s Rule at the FTC

Report from Bloomberg

In Brief – The Republican-led House Government Oversight Committee has begun an investigation of the Federal Trade Commission (FTC) under the leadership of progressive antitrust reformer Lina Khan. The probe claims to focus on issues raised by former Republican Commissioner Christine Wilson, who announced her resignation from the regulatory body in February with a scathing critique of the agency. Wilson’s self-described “noisy exit” accused Khan, her senior staff, and the other Democratic Commissioners, of operational and ethical failings, as well as promoting ideological policies that Wilson claims exceed the legal authority of the FTC. The Committee is asking the agency to turn over a wide range of documents, including an unredacted opinion written by Wilson regarding the FTC’s challenge of Meta acquisition of Within Unlimited, communications among commissioners about the FTC’s authority to clamp down on corporate America, documents about whether Khan should have recused herself from the Meta-Within case, and any related communications from the Biden White House. The FTC is led by a board of five commissioners, three of whom are members of the President’s party, and two of the other party. However, both FTC Republicans have resigned and have not yet been replaced.

Context – Republican support for Khan’s FTC nomination in 2021 was political signal-sending to Big Tech, but her tenure as the agency’s chair has been very divisive. However, neither she, nor the agency she leads, appears to be pulling back. Many contentious FTC undertakings go well beyond “Big Tech”, including new legal thinking on “unfair methods of competition” and an effort to ban employer non-complete clauses, while proposed rulemaking on data privacy and Gig work step in where Congress has chosen not to go. The House Oversight Committee probe joins inquiries by the House Energy & Commerce Committee and the House Judiciary Committee, and look for House Republicans to fight over FTC funding in the appropriations process to highlight business community concerns with agency overreach.

OpenAI CEO Shifts from Complaining About EU AI Act to Talking Up EU HQ

Report from Politico

In Brief – Sam Altman, the CEO of OpenAI, finished up a tour of European capitals by walking back comments that the company might not operate in the EU if the regulatory framework was troublesome enough, and instead discussed how the AI groundbreaker wanted to set up an EU Headquarters unit soon. The AI industry phenom made stops in Spain, France, Poland, Germany and the United Kingdom. At a tech industry event in London, he indicated that the EU’s AI Act, the comprehensive AI legislation that is farthest along among major global markets, could overshoot and make operating in the EU very difficult, in particular due to some changes made by the European Parliament dealing with “generative AI” systems and how copyrighted material is used in “training data”. Altman drew a sharp rebuke from EU Commissioner Thierry Breton “that our rules… cannot be bargained” and was quickly making very conciliatory comments. His focus soon became OpenAI’s interest in setting up an HQ in Europe. While in Paris he said, “If you had to pick just based on the most AI research talent, you’d pick France,” but said that he was “super-impressed by the talent and energy everywhere.” Picking an EU-based headquarters will be important in part because that country’s eventual AI regulator will take the lead role overseeing home-based AI businesses.

Context – Commissioner Breton often says that US-based digital companies need to follow all the rules in Europe, and his list is getting longer, now including the GDPR, Digital Services Act, Digital Markets Act, and most recently the AI Act. Actual effective dates and finalized regulations appear secondary to companies complying with the spirit of the rules. For example, both Breton and Commissioner Vestager have called for “universal rules” governing AI years before the AI Act is in place, describing an AI Pact of “voluntary” rules compliance. It will likely be backed up by Commissioner browbeating similar to what we are seeing with Twitter and the voluntary pact to address online disinformation prior to the DSA going into effect.

Most EU Member States Do Not Support Telecom Operators’ “Fair Share” Plan

Report from Reuters

In Brief – Most of the EU Member State governments are reported to be opposed to the telecom industry’s “fair share” plan to require the largest digital platforms to pay a new fee to help fund the ongoing rollout of 5GF and broadband. Their concerns were recently laid out in a meeting of national telecommunications ministers with Thierry Breton, Europe’s Internal Market Commissioner who oversees the industry and sympathizes with the telecoms’ “sender pays” proposal aimed at video streamers such as Amazon, Apple, Google, Meta, Netflix, and TikTok. At least 18 Member States are reported to have expressed varying degree of opposition to the industry plan, including Austria, Belgium, Czech Republic, Denmark, Finland, Germany, Ireland, Lithuania, Malta, and the Netherlands, while 10 plan backers included France, Greece, Hungary, Italy, Spain, and Cyprus. Critic’ concerns included the lack of evidence of a network investment shortfall, the risk that the fees would be passed onto consumers, and the potential violation of EU “net neutrality” rules.

Context – Telecom companies have long accused large internet platforms of “free riding”. This is despite every internet user paying for the bandwidth they use. These sender pays proposals have often been part of net neutrality debates because telecom companies often argue bandwidth intensive digital platforms should be allowed to pay more to consumer network providers so that their customers can get better quality service, but net neutrality advocates argue that consumers who pay for their internet access will nevertheless not enjoy the same level of service when they use platforms that can’t pay, or don’t pay, the added platform fees. The European Commission opened a formal consultation on the matter in February, and battle lines are forming. Members of Parliament, smaller internet network infrastructure firms, consumer advocates, industry regulators, and now a majority of Member States have expressed concerns. However, the network operators are not about to back away from their campaign.

Meta Agrees to Stop Using Facebook Advertising Data to Benefit FB Marketplace

Report from TechCrunch

In Brief – Meta has offered to stop using advertising data from its core Facebook platform to benefit sales on its Facebook Marketplace classifieds service to resolve an antitrust investigation by the UK Competition and Markers Authority (CMA). The company launched the FB Marketplace in 2016 and integrated it into its core social media platform, which is one of the most widely used digital platforms in the world. FB Marketplace has since grown into one of the top classifieds platforms in many markets, including the UK. The CMA opened an investigation of Facebook in 2021 based on concerns that the company was using data gathered through the massive digital advertising business operating over its social network platform, such as the types of advertisements that a user was clicking on, to direct similar content to the user for goods listed on FB Marketplace in competition with the purchaser of Facebook advertising. Meta is proposing to allow advertisers to opt-out of their advertising data being used to develop FB Marketplace, at least in the UK, to resolve the CMA investigation. The CMA has indicated that it is inclined to accept the offer and is gathering market feedback.

Context – Facebook’s ability to grow the FB Marketplace into a classifieds’ giant using data from its core platform to great benefit is an example of platform “self-preferencing” targeted by advocates of more robust Big Tech antitrust regulation. The European Commission opened a parallel investigation of Facebook and its Marketplace in 2021 and filed its statement of objections last December. Facebook is likely trying to resolve the Commission’s case with similar commitments. But the case is also relevant in the context of the EU’s Digital Markets Act, enacted last year to address conduct like this by “digital gatekeepers” through ongoing regulatory processes rather than case-by-case competition law enforcement. Expect this type of case shifting to DMA regulatory reviews. The UK is slowly moving toward its own standing “Big Tech” digital regulation model with the Digital Markets, Competition and Consumer Bill.

EU Setting a July 6th Deadline to Make Antitrust Decision on Amazon’s iRobot Vacuum Deal

Report from Reuters

In Brief – The European Commission Competition Authority has set July 6th as a target deadline to clear Amazon’s proposed $1.7 billion acquisition of iRobot, the manufacturer of the Roomba robotic vacuum devices, or oppose the deal due to anticompetitive impacts. The Commission, the UK Competition and Markets Authority, and the US Federal Trade Commission, are all formally probing the acquisition, although none of the three have officially moved to block the deal. In February, reports surfaced that the European Commission was continuing to proceed toward a formal investigation that would focus on privacy and data collection, in particular concerns that the robotic vacuums collect sensitive in-home images. The company has responded to such privacy concerns with claims that the devices have only basic room-mapping technology, and a company spokespeople have continued to reiterate that the ecommerce powerhouse is working cooperatively with the relevant regulators in their reviews of the merger.

Context – This deal is another opportunity to see the new era of sequential acquisition reviews with the roulette of US, EU, and UK enforcers. Roomba’s market share slide, the company’s lack of profitability, and no Amazon vacuum raise questions about the strength of a legal challenge. However, Big Tech critics quickly raised a wide range of concerns, including that Amazon could leverage their online commerce dominance to undermine other robot-vacuum competitors and reinforce their overall smart-home product suite. Then there are less traditional privacy-based charges regarding improperly collected in-house images. Similar privacy accusations recently surfaced about in-car videos and Tesla. The US FTC, led by Lina Khan who rose to prominence as an antitrust reformer challenging Amazon’s business model, declined to take on Amazon’s larger MGM and One Medical acquisitions, but continues to probe the iRobot deal and reports indicate that the robot vacuum manufacturer’s residual market share of 62% might make FTC litigators more comfortable.

Vestager Claims US-EU AI Code of Conduct for Companies Possible Very Soon

Report from Reuters

In Brief – European Commissioner Margrethe Vestager believes that a draft code of conduct on artificial intelligence (AI) could be drawn up by the US and EU within weeks and that companies could sign up soon after. Her comments followed a meeting of the US-EU Trade and Technology Council (TTC) that dedicated attention to AI policy in the wake of chatbot phenom ChatGPT. The EU has been developing a comprehensive regulatory plan for AI products and services for five years, with the EU Commission (2021) and EU Council (2022) proposing draft AI Act legislation built on a risk-based framework to regulate AI applications like surveillance, medical devices, and autonomous vehicles rather than underlying technology. But ChatGPT caused a rethink by the EU Parliament, which is calling for new regulatory mandates on “generative AI” and “foundational models”. Vestager, who champions the AI Act but says it will not be fully in place for “two and a half to three years” which she called “way too late”, is also calling for a voluntary company AI Pact in Europe.

Context – The US has approached AI policy with high-level principles and goals. If the US and EU do quickly create some type of company (and lab?) code, will it be more like a quasi-regulatory EU AI Act, or more like US-style principles? Expect the latter. First, look to the recent G-7 leaders meeting. It stayed very high-level with agreement on promoting “trustworthy” AI but no specifics, and Japan, who is leading the new G-7 AI effort, is very supportive of AI work. Then consider that the statement on AI from the TTC meeting leads off with creating an “expert group” on AI “terminology and taxonomy”, reminding us that there is little agreement today on what exactly “AI” even means. Despite the wishes of many US progressives for EU-style tech regulation, that appears a long way off on many fronts. There are reports of deep divides on AI regulation inside the Biden Administration. Finally, progressive digital policy expert and former Biden aide Tim Wu warns that AI licensing and regulating schemes backed by giants like Google and Microsoft could protect their current dominance and dissuade innovators.

Meta Threatens to Block News in California If State Mandates Media Payments

Report from the Wall Street Journal

In Brief – Meta has said that it will block the posting and sharing of news media stories on Facebook and Instagram in California if the state enacts the Journalism Preservation Act. The bill, overwhelmingly passed by the State Assembly, requires large digital platforms such as Facebook and Google to pay news media companies a “journalism usage fee” when news content a media company published appears on the digital platform next to advertising. Meta argues that much of the media content posted on its services is posted by the media companies themselves, who benefit from its platforms’ users and reach, and that media consolidation in the state was occurring well before Facebook emerged as a major digital platform.

Context – Newspapers have been complaining for two decades that the internet ruined their business model as advertisers shifted more and more of their ad budgets to digital platforms. In recent years, big media companies have been lobbying governments around the world to require the largest digital ad platforms, namely Google and Facebook, to pay them more money. Australia broke ground on behalf of media companies. Google and Facebook both drew lines in the sand in Australia that are emerging elsewhere. Meta rejects paying for news content posted by users to the platform, including by media companies themselves. Google opposes being required to pay for simple search links. In Canada, where similar legislation is under consideration, Meta is also threatening to block news. Google has tested a change to its Canadian search engine to remove news covered links from its basic search function. Both digital giants also created curated news services in recent years and paid hundreds of millions of dollars to media companies to reduce political heat. Google has announced expansions of their efforts. Meta, on the other hand, is going the other way. They claim most social media users increasingly don’t want news stories and prefer TikTok-style entertainment. “News” brings misinformation, angst, and strife. So, Meta turning off news in the face of a media company payments mandate might be a real possibility. That would be interesting.

US Supreme Court Gives a Boost to Sec. 230 Defenders in Child Porn Liability Case

Report from Bloomberg

In Brief – The US Supreme Court has chosen not to hear a challenge to a Ninth Circuit of Appeals decision narrowly interpreting the federal Fight Online Sex Trafficking Act (FOSTA), an exemption to Sec. 230 enacted by Congress to better enable victims of online sex trafficking to challenge perpetrators in court. The case pitted victims of child pornography against Reddit, where photos and videos of the plaintiffs were posted on forums. The plaintiffs argued that Reddit’s structure, and its inability to effectively police its platform, allowed the illegal activity to occur and therefore the company could be held liable under FOSTA’s Sec. 230 exemption. However, the appeals court ruled that FOSTA’s “beneficiary” liability standard for a platform like Reddit required it to have “actual knowledge of the trafficking” and that it “knowingly participated in or benefitted from a sex trafficking venture,” which the plaintiffs did not prove.

Context – This decision is an opportunity to do some Supreme Court Sec. 230 tea leaf reading. Like Congress, the justices appear to share general concerns about online activity but not agreement on specifics. Three main legal threads are active: (1) Limiting Sec. 230 to expand platform liability for general bad activity; (2) Restricting how Sec. 230 gives flexibility to digital platforms to police objectionable content, which has led to charges of anti-conservative viewpoint discrimination; and (3) Arguments that platform actions to sort and recommend user-generated content are distinct from content moderation and are not covered by Sec. 230. The view that the Court was primed to narrow Sec. 230 was fueled by comments from some conservative justices, especially Thomas and Alito. But they seem very focused on Thread 2. Twitter v Taamneh deals with Thread 1. Thomas wrote that pro-platform opinion and all nine joined. On Thread 3, the Court chose to pass on using Gonzalez v. Google to address the matter. They are probably looking for a “better” case on a tough issue. The Court will likely need to deal with Thread 2 in the context of the Texas and Florida social media law cases.

Report from the Wall Street Journal

In Brief – Nearly 200 workers in Nairobi, Kenya, who worked for Sama, a California-based contract employee firm, and provided online content moderation services to Meta, are suing both US-based companies in Kenyan court for worker exploitation and unfair termination. Their work included viewing highly disturbing and graphic violence and sexual predation, which they allege caused them mental harm that they claim Sama and Meta did not appropriately minimize or treat. They are also suing to block Sama from terminating them as the firm shuts down its Nairobi-based online content moderation business. Sama denies the allegations, says it supported its workers with health benefits, and has the right to exit the business. Meta is challenging the Kenyan court’s decision to name it as a co-defendant, arguing that it was never the workers’ employer, that it has no employees or offices in the country, and that the Nairobi Labor Court does not have jurisdiction over its business.

Context – The appropriate classification of contractors is an ongoing tech industry issue, and not just Gig-style workers. Many digital companies use contract workers employed by firms like Sama, side-by-side with their own employees. The US NLRB recently deemed Google the co-employer of 50 contract workers employed by Cognizant Technology Solutions in Austin, Texas, who voted to unionize. Of course, Google is challenging the ruling. The plight of social media content moderators again reminds us why tech policy expert Mark Masnick calls it a job that is “impossible to do well”. Meta has invested growing sums, both for personnel and technology, to try to control highly objectionable content globally. Some of the many who do the work suffer harm. The company settled a class action lawsuit by US-based moderators in 2020 for $52 million and has been claiming for nearly a decade that it is training AI to do more of that work better, but AI makes mistakes. As African tech hubs grow, more global digital platforms are tapping local workforces, including for content moderation and AI training, which is leading to unionization drives.

Twitter Considers Pulling Out of “Voluntary” EU Disinformation Code

Report from Politico

In Brief – Twitter is reported to be considering withdrawing from the EU’s Code of Practice on Disinformation, a voluntary code of conduct for digital platforms intended to promote efforts to combat disinformation. The code was initiated in 2018 and updated in 2022. Including Twitter, 34 firms and trade groups have signed on, including Meta, Google, TikTok, and Microsoft. Social media platforms agree to a set of obligations including tracking and reporting on political advertising, efforts to block the monetization of disinformation, cooperating with fact-checkers, and providing better data access to researchers. Last year, the EU enacted the Digital Services Act (DSA), landmark legislation establishing regulatory oversight for a range of online content moderation practices, with especially stringent requirements for the largest digital platforms, which overlaps the Code in some ways. European Commissioner Thierry Breton, who is a leader on EU digital policy and oversees the Commission body that will enforce the DSA for the largest platforms, responded to the Twitter news by tweeting that withdrawing from the voluntary code will not reduce their obligations under the new law, which he said goes into effect August 25. France’s Digital Minister Jean-Noël Barrot added to the discussion by claiming that if Twitter does not follow the law that they will be banned in Europe.

Context – Musk’s defense of “free speech” and criticism of “woke” thinking has upset many progressives and delighted many conservatives. But more relevant to content moderation policy are likely to be his repeated reference to Twitter following “local laws”, which includes the First Amendment in the United States, but means something entirely different in markets like India, Turkey, and yes, the EU. Since he began his effort to acquire Twitter, Musk has repeatedly said that the company would abide by the legal requirements of the DSA. Of course, that is certain to leave things like regulatory discussions and processes, enforcement actions, and related legal challenges, to play out. Rule of Law stuff. Speaking of “voluntary” codes preceding EU legal regimes, top EU Commissioners are calling for a voluntary company “AI Pact” as the bloc legislates its AI Act.

Report from Reuters

In Brief – As expected, Microsoft has formally challenged the UK Competition and Markets Authority (CMA) decision to block its $69 billion acquisition of giant game developer Activision Blizzard. The company’s appeal will be considered by the Competition Appeal Tribunal (CAT), a body that is charged with looking at the legality of the decision-making processes of the CMA, but not the facts of the case or the regulatory decision itself. As such, it is considered a difficult appeal forum. Microsoft’s petition to the CAT states five grounds for appeal, including that the CMA erred in defining the market for the nascent business of cloud gaming, failed to properly consider the constraints of users having the traditional option of playing a game that is installed on your device, and that they failed to account for the long-term commercial agreements made by Microsoft to bring its games to other cloud services for the next 10 years. Microsoft asked for and received an expedited schedule from the CAT to speed up a judicial review process could take up to nine months.

Context – If you are keeping score at home, the US Federal Trade Commission announced in December that it intended to challenge the Activision acquisition in its internal administration court, and the CMA rejected the deal in April. On the other hand, EU antitrust enforcers, who, like the CMA, focused on the emerging cloud gaming market, found Microsoft’s cloud access commitments compelling and cleared the transaction in May. So far 37 countries have approved the deal, with South Korea being the most recent major market regulator to sign off. Although the FTC has recently had federal judges reject some of its more aggressive tech company challenges, the UK CMA has never overturned a merger rejection even following the CAT sending it back for reconsideration. The new era of sequential acquisition reviews, especially with the roulette of the US, EU, and UK, is calling into question established norms of competition policy. Regulators review the same deal and sometimes appear to partner and venue shop in ways that are causing heartburn among business advocates.