Archive – 2023

News & Insights

December 2023

End of the Year News Lightning Round – 4

Meta Rolls Out What They Hope is a DMA-Compliant Threads

Report from The Irish Times

In Brief – Following a delay that Meta attributed to regulatory compliance demands, the company is rolling out its short-form, text-based Threads service in the EU, Iceland, Liechtenstein, Norway, and Switzerland. The service was launched in most markets last July as a part of Instagram. It was held back in Europe based on reported Meta uncertainties with the regulatory requirements of the EU’s Digital Markets Act (DMA) that now governs so-called digital “Gatekeepers”. Elsewhere, Meta is requiring users to have an Instagram account to use Threads. However, in Europe, where regulators object to requiring users to link accounts for different services, a person will be able to create a Threads profile via their Instagram account or use a limited-capability version of Threads without an Instagram-Threads profile.

Context – The DMA is intended to be fully operational in spring 2024. There are six gatekeepers — Amazon, Apple, ByteDance, Google, Meta, and Microsoft – with 22 covered platforms. Instagram is one. There are 18 so-called “Do’s and Don’ts” governing the covered platforms. Creating an alternative Threads option is Meta’s second recent move to comply with European regulations by having a different user experience than globally, following their decision to offer paid, ad-free versions of Facebook and Instagram.

US Government Cutting Back Reports to Big Tech About Election Campaigns

Report from the Washington Post

In Brief – US Government officials have reportedly stopped sharing warnings with Meta and Pinterest regarding foreign-based online information campaigns that appear to be aimed at influencing the 2024 elections. The change is likely linked to the lawsuit filed by the Republican Attorneys General from Missouri and Louisiana arguing that aggressive action by federal officials, including from the Biden White House, FBI, and Department of Homeland Security, pressing platforms to restrict certain posts and online accounts related to highly politicized issues such as COVID and vaccine policies, amounted to unconstitutional censorship. The suit has won favorable rulings in the US 5th Circuit. However, an injunction blocking the White House and some agencies from communicating with the platforms was paused by the Supreme Court in October pending a High Court decision next year.

Context – The High Court will clarify the First Amendment boundaries of when government “jawboning” becomes censorship by proxy. Until then, the biggest social media platforms still have significant internal capabilities to make their own content moderation decisions. There are also countless private sources providing advice, both unsolicited and requested. Media reports such as these, with comments of alarm from government officials, are likely part of a court lobbying campaign.

UK Judges Advised on Use of AI Chatbots to Develop Rulings

Report from Gizmodo

In Brief – The Judicial Office, a branch of the UK Ministry of Justice that supports the judges and courts across England and Wales, has released guidance for the use of artificial intelligence tools, such as Generative AI chatbots, in official court duties. The guidelines do not prohibit their use, but they do attempt to inform judicial users of the limitations and characteristics of the systems, including the reality of authoritative sounding but made-up materials, often called “hallucinations”, as well as the fact that the so-called “training materials” are often US-centric and therefore might not appropriately reflect UK legal tradition. The agency also gives privacy and data security warnings related to the likelihood that data inputted into the services through prompts are often themselves a form of training material and may be used by the systems in future responses to other users.

Context – Lawyers and judges are obviously using AI chatbots. Hopefully, by now, they know that hallucinations are essentially baked into the technology so they must closely check the work. The legal community learned from the experience of the two New York lawyers sanctioned in federal court in June for submitting a brief filled with fabricated cases. The warnings to UK judges using a robot clerk are well taken, especially if they input sensitive material into a general AI system.

End of the Year News Lightning Round – 3

Dem AGs File Supreme Court Brief to Protect Social Media Regulation

Report from WBNG

In Brief – Twenty-two of the 24 states with Democratic Attorneys General have submitted an amicus brief to the Supreme Court urging the court to preserve the right of states to regulate social media platforms when it rules on the First Amendment implications of the laws enacted by Texas and Florida to regulate social media content moderation practices. The Democratic AGs’ brief claims to be “in support of neither party” and that while states have different views on how to regulate social media platforms, they “share a strong interest in preserving their sovereign authority to regulate social media platforms,” citing concerns such as impacts on teen users, drug sales, hate-fueled violence, and sex trafficking.

Context – The High Court will be hearing a set of three cases involving the First Amendment and social media. They have already had oral arguments on whether government officials can block users from engaging with them on their “private” social media feeds. Then comes the Florida and Texas case involving so-called viewpoint discrimination. Finally, there is the question of deciding when communications by government officials to encourage social media content moderation amounts to state censorship. Democrats have consistently opposed the Texas and Florida laws but, consistent with social media regulation in general, they do have their regulatory wish list and are trying to thread a needle on state authority.

Microsoft Greatly Expands “Labor Neutrality” in Deal with AFL-CIO

Report from the New York Times

In Brief – Microsoft has jointly announced with the AFL-CIO that the company will stay neutral if any group of its US-based workers seeks to unionize. The policy expands on a narrow labor neutrality agreement Microsoft reached with the Communications Workers of America in June 2022 involving workers in the video game studios owned by then acquisition-target Activision-Blizzard. Several hundred Microsoft game testers have unionized since. Part of the new agreement between Microsoft and the AFL-CIO includes collaboratively addressing how AI will impact work and employment in the company.

Context – Microsoft’s winning strategy on the Activision deal included a global goodwill campaign aimed at progressive antitrust regulators. The company’s acquiescence to labor organizing was seen in that light, as was their support for major antitrust reforms and app store regulation, and software license changes to promote European cloud services providers. The latest move on labor organizing, which goes beyond any other US-based tech giant, indicates the progressive policy shift may run deeper. The inclusion of how AI will apply to work is especially noteworthy given Microsoft and OpenAI’s role in the field, and the fact that one of its existing videogame tester union contracts governs the use of AI.

Public Support for US TikTok Ban is Falling Away

Report from the Washington Post

In Brief – survey report from Pew Research reveals that public support in the US for banning TikTok has fallen in recent months, from 50% in March to 38% in October. Although TikTok criticism in political circles has been bipartisan, and both the Trump and Biden Administrations have called on Chinese-based ByteDance to sell its US TikTok operations, Republican officials have been more aggressive. The Republican-dominated Montana state government enacted a first-in-the-nation ban on the app in May, which has since been blocked by a federal judge. Among Republicans, support for banning the app has fallen from 60% to 50%, while for Democrats support for a ban has fallen from 43% to 29%.

Context – Public sentiment and politics aside, when the Trump Administration tried to ban Chinese-based TikTok and WeChat in 2020, federal judges were very sympathetic to lawsuits challenging the app bans. TikTok won an injunction, as did a TikTok content creator. WeChat users even won an injunction on First Amendment grounds, and WeChat is a recognized part of the Chinese internet censorship regime. TikTok says they are not. If CFIUS, a Cold War era regulatory body, tries to force a sale of TikTok, it would kick off a legal battle over the limits of the body’s legislative backing in the post-Cold War environment.

End of the Year News Lightning Round – 2

Apple Looks to Settle Tap-and-Go Payments Inquiry in Europe

Report from Reuters

In Brief – Apple has reportedly offered to allow rival payments services to access the Near-Field Communication (NFC) technology on the iPhone so that they can add so-called tap-and-go mobile payments to their mobile wallet services. The iPhone giant, who has blocked direct competitors to the Apple Pay service from using NFC to create a similar user experience, has long argued that they restrict third-party apps from integrating with NFC to protect user privacy and enhance security. They reportedly hope their latest offer can settle the antitrust complaint brought by the European Commission in May of 2022. The Commission is likely to seek feedback over the next month from stakeholders including payments industry rivals, users, and other customers such as banks and retailers.

Context – This EU antitrust complaint about Apple payments but is NOT part of the massive fight over “in-app payments” hounding Apple and Google (who just lost a jury trial in US federal court to Epic Games). Those are about app developer fees. This is one of many complaints alleging that Apple discriminates against third-party apps that compete with their services. It is more like Spotify’s claim of discrimination that is expected to bring a similar EU complaint. Finally, Apple is a “Gatekeeper” under the EU Digital Markets Act and it is expected to address this kind of conduct when up and running in the spring.

More AI Industry Funding for Expert “Fellows” Serving as Hill Staff

Report from Politico

In Brief – The American Association for the Advancement of Science (AAAS), a venerable science non-profit established in 1848, has created a “rapid response cohort” of congressional fellows to aid key Senate offices in the development of AI legislation and policy. The effort is backed by substantial contributions from top AI companies including Microsoft, OpenAI, Google, IBM and Nvidia. Former Microsoft executive Crain Mundie, who still advises MS and OpenAI, takes credit for encouraging the AAAS, who has operated a government fellow program for 50 years, to target fellows on AI policy. The initial AAAS AI fellows are working for Sens. Martin Heinrich (D-NM), Mike Rounds (R-SD), Ron Wyden (D-OR), Bill Cassidy (R-LA) and Mark Kelly (D-AZ.).

Context – When news broke back in October that Open Philanthropy, a non-profit organization backed by Silicon Valley billionaires, was funding more than two dozen “fellows” to work in key congressional offices, federal agencies, and think tanks on AI and Biosecurity policy, it raised eyebrows even among PEI readers with deep experience and knowledge of the dark arts of policy influence. Is it already old news? On the other hand, a recent report that large tech companies are major contributors to top universities and research professors on technology and policy is the opposite. Yawn. That’s been going on forever.

Amazon to Defend iRobot Acquisition in EU Commission Hearing

Report from Reuters

In Brief – Amazon is defending its $1.4 billion acquisition of iRobot, the maker of Roomba robot vacuums, in a late December closed-door hearing of European Commission competition officials and national antitrust regulators. EU officials began a formal investigation in July and followed up with a Statement of Objections in late November, including specific concerns on strategies they believe Amazon could pursue to undermine robot vacuum rivals who sell on the massive Amazon marketplace. Commissioner Didier Reynders has said he wants Amazon commitments to fair treatment for rivals.

Context – Despite the relatively small price tag, this acquisition is interesting on a few levels. First, the UK CMA already cleared it. Alone, that’s not surprising, but it was interesting to see that the CMA understands that Amazon’s most profitable ecommerce strategy often involves sales by third-parties who buy Amazon logistics and on-site advertising. Those commissions often reach 50%, far above first-party retail sales. This is also why Amazon’s recent antitrust settlements in the EU and UK try to limit Amazon’s ability to preference sellers who purchase logistics services. Does the Commission question whether that settlement will work? Or, if it does, do they suspect Amazon will preference their own? And how about the DMA?

End of the Year News Lightning Round – Part 1

Platform Labor Directive Deal in Europe

Report from TechCrunch

In Brief – After months of negotiations, the European Council and Parliament have reached high level agreement on the Digital Labor Platforms Directive to regulate Gig-style work platforms in the EU. The Commission proposed an initial draft in December 2021. A top goal has been to standardize when platform workers should be classified as employees of the platform, with applicable employee rights and benefits, rather than as independent contractors. The agreement sets five conditions of platform control over work, and if two are met then there is a rebuttable presumption that the workers are platform employees. The directive also sets new rules to govern automated monitoring and algorithmic decision making related to workers.

Context – As the EU has been plodding forward trying to find a balance between ending so-called “phony” independent contractors while enabling new work models that give skilled freelancers valuable flexibility and independence, labor-backed worker classification initiatives have been stymied in the US since California voters exempted Gig-drivers from state law AB 5 late 2020. In the UK, the High Court recently confirmed that the level of platform control is key there too, with Deliveroo drivers adjudged to be contractors while Uber drivers are workers.

Zulily Marketplace Sues Amazon for Anticompetitive Practices

Report from Reuters

In Brief – Zulily, a nearly defunct ecommerce marketplace that attempted to challenge Amazon with sellers offering lower prices, has filed an antitrust lawsuit against Amazon arguing that the ecommerce and logistics giant bullied third-party sellers to use its high-priced logistics services, forcing sellers to set higher prices on Amazon, while also pushing them to not sell their products for lower prices on other low-cost platforms. Zulily contends that the combination of conditions Amazon imposed on sellers trying to succeed on Amazon’s giant marketplace left few sellers willing to make lower price offers on lower-fee Zulily.

Context – This lawsuit is focuses on the same set of Amazon practices targeted in lawsuits by consumers, the AG of California, and the Federal Trade Commission. While Amazon emerged as a traditional retailer, third-party sellers are now their biggest and most profitable ecommerce business. Its seller fees often reach 50 percent. Plaintiffs contend that many sellers would offer consumers lower prices on lower fee marketplaces, but doing so gets them penalized by Amazon, and they won’t forgo their sales on Amazon to potentially make some sales on a small platform like Zulily. Amazon recently settled antitrust investigations in the EU and UK that aim to stop them from unfairly preferencing sellers who buy Amazon logistics.

Cybersecurity Agencies Agree on Safe AI Practices

Report from Reuters

In Brief – Cybersecurity agencies from 18 countries, including the US, UK, Japan, and Germany have agreed on policies and practices that they recommend companies designing and employing AI should use to keep customers and the wider public safe from misuse of their systems. The Guidelines for Secure AI System Development are not proposed as mandates or regulations. Instead, they are largely high-level recommendations such as monitoring AI systems for abuse, protecting data from tampering, and vetting software suppliers. The signatories include representatives of the so-called Five Eyes, who are the US’s top security and surveillance partners, eight European countries, Japan, South Korea, and Singapore from Asia, Israel, with its world class cybersecurity industry, Chile, and Nigeria.

Context – The most newsworthy AI public policy development of the year was the EU appearing to wrap up agreement on their AI Act creating a comprehensive regulatory structure for all manner of applications and technologies. As the effort heated up, a big question was whether the EU would be a global trailblazer or outlier. Even after the “final” deal was struck, French President Emmanuel Macron was expressing concerns of over-regulation. Although not as exciting, the AI guidelines for cyber security are likely to be more reflective of where most governments want to go on AI policy at this point.

UK CMA Investigating Antitrust Implications of Microsoft Investments in OpenAI

Report from Reuters

In Brief – The UK Competition and Markets Authority (CMA) has begun the first stage of what may become a merger review of Microsoft’s massive investments in OpenAI. The regulator has initiated an open comment period inviting “all interested parties” to share their views on the nature and implications of the relationship between the digital giant, which is the second highest valued company in the world, and the AI phenom. The CMA’s inquiry is focused on whether Microsoft’s investment and partnership involves a level of control that is akin to a merger or acquisition. The relationship between Microsoft and OpenAI has been widely reported since the startup broke into the public consciousness by releasing ChatGPT over a year ago, including Microsoft investing $10 billion after ChatGPT exploded on the scene, which is said to confer 49% ownership of the for-profit part of OpenAI, as well as integrating OpenAI technology into major Microsoft services. However, the role of Microsoft in last month’s internal drama at OpenAI was also noteworthy, with deposed CEO Sam Altman responding by quickly jumping into a top AI development role at Microsoft with many OpenAI employees threatened to go with him. When the OpenAI Board doubled back and restored Altman, they reconstituted the OpenAI Board absent the leaders who opposed rapid AI commercialization and added a Microsoft representative in a non-voting capacity.

Context – You often hear two refrains from advocates of regulating artificial intelligence. One is to not make the “same mistakes” with AI that were allegedly made with social media, apparently meaning that government should step in early to regulate. The second is that AI is quickly becoming dominated by the tech giants. The emergence of OpenAI as the leader in LLM chatbots somewhat belies the concern, as it was a startup that outperformed giants like Google, Microsoft, Apple, Amazon, and IBM. With facial recognition AI, unknown Clearview AI beat the giants to a functional service. And AI was reportedly key to TikTok content algorithms that led to their massive global social media growth. All counter narrative.

European Commission Not Likely to Pull Apple’s iMessage into the DMA

Report from Bloomberg

In Brief – People familiar with the European Commission’s in-depth review of whether Apple’s messaging service iMessage should be regulated as a Core Platform Service under the Digital Markets Act (DMA) report that the regulators are leaning against that designation. For messaging services, key obligations are interoperability and the ability to connect seamlessly with competitors’ messaging services. Apple’s iMessage has proudly never done so, with the company’s “blue bubble” texts and many features only operative when all participants in a message thread are using Apple devices. Meta’s Facebook Messenger and WhatsApp were included in the initial release, but the Commission said that it was further studying Apple’s iMessage with a decision by February. Google and its telecom allies have argued that Apple’s service meets the qualitative thresholds of the DMA and that applying the rules would benefit European consumers and businesses, but the Commission reportedly is finding that too few businesses use iMessage to meet the law’s thresholds.

Context – The DMA uses proactive regulation in place of traditional antitrust enforcement to police the most popular platforms of the largest digital companies. In September, the Commission announced that Amazon, Apple, ByteDance, Google, Meta, and Microsoft were gatekeepers with a total of 22 core platform services. Google, with eight, has the most. ByteDance has just one, TikTok, and the Chinese company is challenging that designation in court, arguing that TikTok is a pro-competitive challenger to established and dominant social media and video platforms. Meta is appealing the separate designation of its Marketplace and Messenger services, arguing they should be regulated as part of the Facebook platform. In related news, Apple has announced that it plans to upgrade the ability of iMessage to work with outside services, including adding the ability to accept high-resolution photos and video, although it will not be a full integration with RCS messaging and will not end the so-called “green bubble” issue.

FTC Asks Federal Court of Appeals to Overturn Loss on Microsoft-Activision Deal

Report from CNN

In Brief – Lawyers from the Federal Trade Commission (FTC) and Microsoft squared off in front of the Federal 9th Circuit Court of Appeals as the regulator asked the court to overturn Federal Judge Jacqueline Scott Corley decision to reject the agency’s motion to block Microsoft’s $69 billion deal to buy videogame giant Activision-Blizzard while it challenged the acquisition in its internal court system. In her opinion, Corley said that the FTC had failed to show that Microsoft owning Activision games would harm competition in the videogame console or cloud-gaming markets, and even indicated that the deal would more likely lead to “more consumer access to Call of Duty and other Activision content.” The FTC argued that the judge had been too deferential to Microsoft’s promises to allow other platforms to access popular Activision video games, including through a series of license agreements, and held the agency to too high a standard of proof that Microsoft planned anticompetitive harms. Microsoft responded that Corley’s ruling made “clear factual findings” regarding the expanded access to Activision games brought about by the acquisition. The acquisition was completed by Microsoft on October 13 following approval by the European Commission in May, Judge Corley’s ruling against the FTC in July, and lastly the UK Competition and Markets Commission approving a revised deal that expanded access to Activision games on competitor cloud game systems.

Context – The FTC continuing to battle away in federal court to block the merger is entirely consistent with the Biden Administration’s public commitment to revitalizing antitrust enforcement and pushing back against consolidation in tech markets and more broadly. Less than a year before the 2024 presidential election, there is little reason to expect anything less than full throated backing for an agenda that energizes progressive activists, as even (expected) court losses can be used to argue for legislative reforms that have long been on their wish list.

Amazon Files Motion to Dismiss Landmark FTC Antitrust Case

Report from Bloomberg

In Brief – Amazon has filed a motion to dismiss the Federal Trade Commission’s major antitrust complaint. The ecommerce and logistics giant argues that requiring sellers to offer their lowest price on Amazon, and to buy Amazon’s logistics services to provide delivery, is pro-competitive.

Context – Amazon is a complex conglomerate. Some businesses are digital (AWS, Twitch). Some are crossovers between hardware and digital (Ring, Alexa, Kindle). Some are loss-leaders to drive Prime subscriptions (Prime Video). Their OG first-party retail business is giant but has low margins. The third-party seller marketplace services business is even larger and has high margins. Their unique and massive logistics business is their biggest moat, has the most employees, and includes huge gig work platforms too. Regulators have had a tough time focusing given the kaleidoscope of interconnected services. Recently, the relationship between Amazon’s logistics business, the treatment of third-party sellers on their marketplace, and how Amazon’s algorithms influence who makes sales, is getting the most attention. This is because Amazon commissions imposed on third-party sellers often approach 50% because of logistics fees, and Amazon algorithms reward sellers who buy logistics to the point that most sellers recognize it is necessary to make meaningful sales. In the EU and UK, Amazon has settled antitrust cases promising to break that link. In the US, the same issue has led to antitrust complaints alleging that Amazon unfairly preferences sellers who buy Amazon logistics (the same conduct targeted in Europe) and penalizes sellers who offer lower prices elsewhere, even when lower logistics costs would allow lower prices. Those “price fixing” complaints are largely contingent on Amazon being such a dominant marketplace platform for small online retailers that sellers cannot afford to forgo sales on Amazon, and instead raise prices elsewhere online to protect sales on Amazon. A private antitrust suit and one from the California AG have survived Amazon’s motions to dismiss. We’ll see about the FTC’s complaint.

Epic’s Big Jury Trial Win Over Google Is Opposite of Judge’s Apple Ruling

Report from the Wall Street Journal

In Brief – A jury in Northern California has found Google guilty of violating federal competition laws in the antitrust case brought by videogame giant Epic Games. The panel, which deliberated for a few hours, determined that Google illegally ties the Google Play app store and its Google Play Billing payment services, and that some distribution agreements and deals Google makes with app developers and phone manufacturers were anticompetitive. When Epic’s similar lawsuit targeting Apple was decided by a federal judge in 2021, the iPhone maker was determined not to violate federal antitrust law. Google says it will challenge the verdict, reiterating that Android competes fiercely with Apple and that its rules and commissions are needed to operate a competitive alternative that benefits consumers.

Context – Apple’s closed “walled garden” ecosystem that earns more app revenues than Google’s Android does not violate federal antitrust law. While it seemed counterintuitive that the more open system, which earns less money, could be found violating the same antitrust laws, we’ve speculated for months about how it could. First, Apple’s system has been closed from the beginning. No confusion and clear messaging to everyone, especially consumers. And Apple makes the phones. No messy deals with phone makers. Android is a complex and confusing hybrid. The base code is “open source”, but highly restrictive rules apply to manufacturers wanting to brand Android devices, and complex business deals are struck to preference Google apps. They are causing big regulatory and legal problems. Plus, Google had a jury trial. Did they really think a jury trial was a good idea in such a complex case? Finally, Google engaged in conduct involving deleting case-related internal message threads after Epic’s suit was filed and employees had legal “hold orders”. Presiding Judge James Donato was very upset about the conduct and the jury received instructions to believe the deleted communications may have indicated guilt. Donato will now determine the remedies and the issue is also surfacing in Google’s other antitrust cases.

TikTok Investing in Indonesia’s Tokopedia After Government Bans TikTok Shopping

Report from Bloomberg

In Brief – In a bid to revive its Indonesian ecommerce aspirations following a government ban, TikTok has announced that it will invest in Tokopedia, the online marketplace unit of Indonesia’s GoTo Group, and the pair will jointly develop a new online shopping service in the country that will operate separately from both companies’ core platforms. Indonesia was the first and largest market for TikTok Shop, a live-streaming and video-based ecommerce marketplace integrated into the core TikTok app. It was launched in 2021 and quickly grew into the country’s fifth-largest ecommerce platform, disrupting Indonesian retail both online and from traditional small shops. In response, the Indonesian Government imposed a crash regulatory change in October aimed at TikTok, prohibiting ecommerce transactions from occurring over social media platforms. TikTok, a wholly-owned subsidiary of Chinese-based ByteDance, duly committed to complying with local laws and regulations and shut down TikTok Shop Indonesia. Tokopedia is Indonesia’s second largest ecommerce marketplace, and the companies said that the tie-up aimed at bringing a new ecommerce platform to the market would be contingent on regulatory approval. The third largest ecommerce marketplace in Indonesia is Lazada, a Philippines-based business owned by Chinese-based Alibaba.

Context – TikTok’s massive global popularity, expanding business aspirations, and an exploding user base that is uploading content that stretches far beyond the platform’s once traditional base of short music and dance-based videos, is leading to a diverse range of regulatory challenges. Of course, there are concerns over data security and censorship due to Chinese influence, most notably the ongoing standoff in the US, but geopolitical concerns also led to the app being shut down in India in 2020. The platform’s emergence as a major ecommerce disrupter in Indonesia led to a shopping shutdown, and Malaysia is reported to be considering something similar. Finally, adult-themed live-streaming on TikTok is reportedly a factor in efforts to ban the platform in Kenya and Nepal.

EU Wraps Up AI Act to Ensure They Play the Role of Lead Tech Regulators

Report from DW

In Brief – European Union negotiators have agreed on the parameters of the AI Act, the first legislation to set up a comprehensive regulatory framework for artificial intelligence in a major Western market. The deal largely sets out tiered, risk-based rules for AI applications as proposed by the European Commission in 2021, with the addition of regulations for “foundation models” that was included by the Parliament following the release of Chat-GPT. AI-enabled products and services are defined as unacceptable risk, meaning prohibited (such as social credit scoring or real time biometric tracking), high risk, which will require pre-approval from regulators based on safety parameters (such as self-driving cars or medical machines), and limited risk applications given transparency and testing requirements (such as touching up photos or suggesting videos). The compromise that resolved the final hurdle was an exemption allowing state security services to use otherwise prohibited biometric tracking in some instances, such as identifying violent criminals. Generative AI foundation models are regulated, with the largest models, based on computing power, considered a “systemic” risk. They must inform regulators how they work, were trained, the data sets involved, and how they prevent illegal content from being generated. The compromise that was reported to have won over France and Germany was to impose the toughest regulations only to the largest models, which currently only covers the latest from OpenAI, and largely exempt “open source” models, which freed the top AI companies from France and Germany. If the bill’s final language is wrapped up in the coming weeks and passed by the Parliament and Council next spring, it goes into effect in 2025. But French President Emmanuel Macron is already expressing grave concerns.

Context – If it all goes through, this is a huge deal. But don’t expect a wave of followers. The EU may proceed towards regulators overseeing AI systems ex ante, theoretically approving or denying their operations. The rest of the world, especially the US, UK, and Japan, does not seem to be following and may choose to appear more friendly to AI investment and entrepreneurs, waiting to regulate AI systems on a case-by-case or specific use-case basis.

Amazon’s Twitch to Shut Down in South Korea Over Telecom Usage Fees

Report from Yonhap News Agency

In Brief – Twitch, a popular live video streaming service owned by Amazon that is widely used by gamers to watch and perform live play-throughs and commentary, has announced that it plans to shut down its South Korean service in February due to high network usage fees that it must pay telecommunications companies. South Korea is a very large market for many video games and “esports” of competitive game playing, with top players earning national followings and major salaries. Top Twitch executive Dan Clancy, who announced the move on a corporate blog, said that operating Twitch in South Korea was “prohibitively expensive” with network usage fees 10 times greater than in other countries, and that efforts to adjust the video quality on the service to reduce bandwidth costs were not successful. He stressed that Twitch has been operating at a significant loss in the country and sees “no pathway forward for our business to run more sustainably” in South Korea.

Context – Telecom executives in markets globally have long called for digital platforms to pay their “fair share” of the cost to update and maintain broadband networks. The EU has been front and center recently. Many are envious of industry colleagues in South Korea, the one major market with a longstanding regime of “data usage fees” requiring digital content companies to pay network providers on top of the fees consumers pay for their broadband service. Several US-based companies, including Netflix and Facebook, have pushed back against the charges with limited legal success. Legislation to clarify that non-South Korean online companies also owe the fees has been tied up in the National Assembly based on divided consumer opinions. Netflix and SK Broadband, one of South Korea’s largest ISPs, recently announced that they are ending their long-running legal battle over the fees and are instead are creating a strategic partnership where the streamer would will pay for a range of service enhancements from the network company. Expect concerns over deals like that to be part of net neutrality discussions at the US FCC next year.

Marathon EU AI Act Negotiations Stretch Into Friday As Backers Hope to Close Deal

Report from Reuters

In Brief – The December 6th AI Act “trilogue” negotiating session that backers hoped would see the European Parliament, Council, and Commission wrap up their differences and agree on a comprehensive legal regime for AI-enabled products and services stretched into Friday the 8th in a marathon effort to wrap up the five-year effort. As expected, two substantive disagreements were at the forefront. The Parliament supports applying a meaningful regulatory scheme to so-called “foundation models” like ChatGPT, while national governments are divided, with France and Germany in particular pressing for the AI Act’s most intrusive risk-based regulation being applied to specific AI applications and services rather than the underlying technology. The second major hurdle is a longstanding divide between the Parliament and the Council over member states giving law enforcement and security services flexibility to use AI-enabled biometric surveillance, including facial recognition. Reports indicate that a compromise on foundation models might be reached, possibly exempting “open source” AI models from the strictest regulations that would apply to proprietary systems from current AI leaders. The Council is reported to hope their deal on foundation models would spur a surveillance compromise from Parliament.

Context – The push to close the AI Act is energized by the view that if talks slip into next year, then they won’t wrap up before the EU Parliament’s elections. Champions of Europe as the global leader on tech regulation see that slipping away on AI, the next big thing. They should relax. Nobody else is seriously legislating. The most interesting development is the reported compromise on foundation models that exempts open source versions. France and Germany emerging as strong opponents of regulating foundation models, which has caused such trouble, is attributed to concerns over the impact on emerging firms Mistral from France and Aleph Alpha from Germany. They both follow open-source models. Most of the biggest US-based companies don’t, although Meta is an exception. A workaround to carve out the top EU firms? I’m shocked, shocked.

Five Social Media CEOs to Appear at January Hearing of the Senate Judiciary Committee

Report from The Verge

In Brief – CEOs of five digital platforms often accused of harming teen users have agreed to testify at a January 31 hearing of the Senate Judiciary Committee. The hearing on child exploitation online will feature Meta CEO Mark Zuckerberg, TikTok CEO Shou Zi Chew, X’s CEO Linda Yaccarino, Snap CEO Evan Spiegel, and Discord CEO Jason Citron. The committee initially aimed to bring in the CEOs of X, Snap, and Discord for a December 6 hearing, but have rescheduled with all five next January. Zuckerberg will be testifying before Congress for the eighth time, while TikTok’s Chew testified in front of the House Energy and Commerce Committee in March. While senior executives for X (when named Twitter) and Snap have testified on Capitol Hill before, this will be the first appearance of Yaccarino and Spiegel. Jason Citron is the first Discord official to appear as a congressional witness.

Context – Back in July 2020, the CEOs of Amazon, Apple, Google, and Facebook testified at a blockbuster House Antitrust Subcommittee hearing (recap and video here), kicking into high gear legislative efforts to reform antitrust rules for Big Tech that continued through December 2022 and only receded when Republicans took over as the House majority. January’s five CEO Senate hearing illustrates how the issue of regulating social media platform dealings with teen users has emerged as the top tech reform prospect in this Congress. During a recent Senate Judiciary hearing focused on disagreements within Instagram about protecting teen users from dangerous, objectionable, and offensive communications, Senators were uniformly critical of Meta and Big Tech companies, with many calling for legislation to regulate social media platforms and reform Sec. 230. Expect more of the same. Meanwhile, as Congress debates the problem, and some states enact age-based internet regulations (with varying degrees of success withstanding judicial scrutiny), federal lawsuits brought by State AGs, school districts, and private plaintiffs are emerging as the most significant risk to the platforms.

Federal Judge Dismisses Private Antitrust Case Against Google Maps Policies

Report from Bloomberg

In Brief – Federal Judge Richard Seeborg has granted Google’s motion to dismiss the antitrust complaint of digital marketing industry plaintiffs led by Dream Big Media that alleges Google illegally ties its various maps-related services to drive up the cost of digital mapping services generally. This is the second time in 18 months that a federal judge has dismissed Dream Big Media’s complaint. Judge Seeborg argued that the complaint was severely weakened by the fact that even the plaintiffs admit that several competitors offer mapping services that are “as good as or better in terms of performance and/or cost” than Google.

Context – Despite the string of losses by the private complainants in US court, Google’s Maps is an increasingly popular target of antitrust scrutiny. The US Department of Justice’s Antitrust Division, led by a long-time Google nemesis, is actively considering targeting Google’s Maps practices, which would add to the two major federal antitrust cases, on search and advertising, already underway. The two dismissals issued by federal judges must be informing their strategy. In Europe, Google Maps is one of seven Google services designated as “Core Platform Services” under the EU’s Digital Markets Act (DMA), along with Google’s Ad platform, Search, Android, the Chrome browser, YouTube, the Play Store, and Shopping. That is seven of the 22 covered platforms offered by the six “Gatekeepers”. The DMA applies 18 so-called “Do’s and Don’ts”, competition policy-inspired regulatory mandates, to the covered platforms with the European Commission acting as an ex-ante regulator rather than relying on traditional ex-poste enforcement actions. Then there are the Member State competition regulators. Germany’s FCO issued a statement of objections regarding Google’s Maps Platform in June that parallels the DoJ investigation, and the Italian competition authority fined Google in 2021 for anticompetitive practices on their auto services platform. Google Maps may be instructive on how EU competition investigations involving core digital platforms proceed in the context of the DMA.

European Commission Market Testing Microsoft Offer to Unbundle Teams from Office

Report from Reuters

In Brief – EU antitrust officials are market testing Microsoft’s offer to unbundle its Teams chat and video app from its Office 365 product, asking rivals if the plan addresses their concerns. Microsoft announced its Teams offer in August in a bid to resolve a European Commission antitrust investigation that was spurred by a complaint from Slack, a rival business chat and collaboration app owned by Salesforce. Along with potentially resolving the standoff, responses to the Commission’s questionnaire could alternatively help antitrust officials develop a statement of objections that could be sent to Microsoft early next year. On pricing, Microsoft proposed that Office without Teams would be 2 euros per month cheaper than the version with Teams, while offering new Teams customers 5 euros a month for a standalone version, a discrepancy that frustrated some competitors.  Along with concerns over pricing and limited eligibility requirements, the Commission wanted to know from potential customers how well rival’s services operate in tandem with Microsoft’s services, whether it was easy for customers to switch, if network effects potentially block competitors from entering the market, and the future impact of cloud offerings.

Context – Microsoft’s regulatory good behavior campaign and Activision licensing deals helped win over the European Commission on its massive videogame acquisition. Now the company needs to manage a collection of other competition policy issues in Europe with a Commission that might be interested in proving it can be tough on the giant. Along with the Teams investigation, there are complaints from some European cloud services providers that Microsoft uses software licensing to pull EU customers into its cloud infrastructure, and NextCloud has complained that Microsoft unfairly bundles its OneDrive cloud system with Windows. Microsoft’s Windows operating system and LinkedIn have been designated gatekeeper core platforms under the Digital Markets Act regulatory scheme, and the Commission is studying adding Microsoft’s Bing, Edge, and the Microsoft Advertising Platform to the roster as well.

Leading Online Platforms Sign onto Voluntary Charter to Fight Fraud Better in the UK

Report from the BBC

In Brief – With online frauds, including phony investment schemes and romance scams, growing as a high-profile public concern in the UK in recent years, the Sunak Government has engineered a voluntary charter adopted by a dozen top online platforms, including giants such as Amazon, Google, Meta, Microsoft, and TikTok, to better tackle frauds on their platforms. The Online Fraud Charter follows similar efforts in retail banking, telecoms, and accountancy. While the companies signing onto the voluntary Charter are acknowledging that frauds perpetrated using their platforms are harming the public and committing to a range of high-level responses to address the problem, the effort is underway in the shadow of the recently enacted Online Safety Act. That measure, which sets up Ofcom as the new content moderation regulator for digital platforms, has begun its route towards full implementation via rulemaking that will likely back up general commitments from the new Charter with more concrete and enforceable mandates that focus on the most tangible risks facing each company’s unique business model.

Context – One of the changes made to the Online Safety Act in mid-2023 to wrap up years of wrangling over the scope of the bill was to pare back provisions requiring digital platforms to pay financial compensation to victims of scams and frauds that targeted consumers through online ads and messages. The change was driven by concerns over the negative financial impact on the tech sector and potential damage to efforts to promote the country as a hub of digital innovation and investment. As the fraud liability provisions in the bill were pared back, the Sunak Government announced a National Fraud Strategy promoting greater cooperation between government agencies, law enforcement, and the private sector, including major tech companies pledging to take a more proactive approach to combating digital frauds. Hence the Online Fraud Charter. And the company-by-company regulatory schemes via the Online Safety Act will very likely mandate more concrete actions in due course.

AI Regulation Backers Looking for EU AI Act “Trilogue” Breakthrough Today

Report from AP

In Brief – The legislative prospects for the EU AI Act, a five-year effort to create a comprehensive legal regime for AI-enabled products and services, faces a key “trilogue” negotiating session on December 6 where leaders from the EU Parliament, Council, and Commission will try to agree on a unified version. The top two disagreements are clear, whether a strict regulatory scheme will be applied to so-called “foundation models”, and whether real limits on biometric surveillance, including facial recognition, will be applied to law enforcement and security services. When the Parliament enacted its version of the AI Act in the first half of 2023, it added regulation of generalized foundation models in a way that fundamentally changed the AI Act. On facial recognition by security services, the Parliament, pushed by privacy advocates, has consistently opposed the position of the member states to largely exempt security and law enforcement from the law’s limitations. If compromises on these two major issues are not reached this week, many believe that negotiations may slip past next year’s Parliament elections.

Context – Avoid the hyperbole about Europe failing to enact the “gold standard” in AI regulation if they don’t wrap up the AI Act now. Understand three things. First, the AI Act proposed by the Commission in 2021 and adopted by the Council in 2022 are fundamentally different from Parliament’s version. The original was a risk-based regulatory system for enterprises using AI-enabled applications based on the type of service, not underlying technology. Parliament added major regulation of general AI systems. That big shift led to concerns of over-regulation from unlikely EU voices like Macron and Jourova and major EU startups. Germany and France don’t carry water for “Big Tech”. Second, divisions over allowing state security to use AI surveillance are deep and stretch back years. Third, the idea that the US, UK, China, G7, or the UN, might jump ahead with meaningful AI regulation is far-fetched. Everyone is going slowly and advocating cooperative business-government safe AI principles and behaviors. (OK, China’s AI regulation is protecting their online censorship regime.)

Google Agrees to Pay C$100 MN Annually to Make Canadian Media Problem Go Away

Report from the CBC

In Brief – Google has reached an agreement with the Canadian Government to pay C$100 million ($74 mn USD) to comply with the Online Media Act that requires the largest digital platforms to pay Canadian media companies when news content appears on their platforms. The law, passed in June, targeted Google and Meta. Both companies objected to paying media companies for content they don’t actively place on their own platforms. Meta argued they should not be asked to pay when news content was posted by users, especially by media companies themselves. Google objected to paying for links surfaced in basic search results. Both said they would block covered news content rather than make mandatory payments. Meta instituted that policy in August, claiming they wanted Facebook and Instagram algorithms to be accurate when the law went into effect on December 19, and has faced ongoing criticism from government and media officials. Google negotiated with the government, increasingly focused on avoiding the prospect of uncapped payments. The deal sets Google’s payment at C$100 million, below the initial government estimate of C$172 million annually. Google will contribute to a collective fund rather than determine individual payments to individual Canadian media enterprises.

Context – For six months, Canada has been ground zero in the global campaign by media companies to have governments push the largest digital platforms to pay them. The most interesting development is the divergence between Google and Meta, with Meta being increasingly straightforward that they will block news rather than pay a government-set rate for media content they don’t themselves place on Facebook and Instagram. They say big media companies overvalue their content and undervalue Meta’s free distribution platforms. The company might even believe that “news” is a net loser even when “free”. Global observers will be calculating next steps, including in the US Congress and Sacramento. With Google willing to pay, but not Meta, will other large platforms such as X (formerly Twitter) be pulled in?

Federal Judge Blocks Montana TikTok Ban Pending Trial (As Expected)

Report from the New York Times

In Brief – In a ruling that was not unexpected based on the hearing record, US District Judge Donald Molloy issued a preliminary injunction blocking Montana’s first in the nation state law banning downloads of TikTok from going into effect in January. Molloy summed up his decision saying the law “likely violates the First Amendment.” TikTok, as well as a handful of TikTok users in the state, filed suit to block the law following its enactment earlier this year. Based on this decision, the law will not go into effect pending the results of a bench trial in federal court that has not yet been scheduled. The judge argued that the Montana law deprives some citizens of communicating via their preferred means of speech, requiring a high degree of scrutiny under the First Amendment, and that TikTok’s side had better arguments and had demonstrated that it was likely to prevail in the eventual trial.

Context – As we have said from the beginning of Montana’s effort to ban TikTok, their law faces high legal obstacles. When the Trump Administration tried to ban Chinese-based TikTok and WeChat in 2020, federal judges were very sympathetic to lawsuits challenging the app bans. TikTok won an injunction, as did a TikTok content creator. WeChat users even won an injunction largely on First Amendment grounds, and WeChat is a recognized part of the Chinese internet censorship regime. TikTok insists they are not. A federal law banning TikTok would face a high First Amendment bar, and the Cold War-era Berman Amendments limit some other federal government legal avenues. And given how much of the justification to block TikTok is based on foreign policy and national security concerns, a state law is even harder to defend. The years long CFIUS review of TikTok, which could try to unwind a 2017 acquisition that created the current service, probably remains the biggest threat in the US, but if CFIUS eventually orders the company to be broken up, it would kick off an unprecedented legal battle over the limits of the body’s legislative backing in the post-Cold War environment. But Nepal recently banned TikTok.

Max Schrems’ NYOB Challenges Meta’s GDPR Subscription Plan

Report from TechCrunch

In Brief – As expected since Meta first announced their intention to offer paid, ad-free versions of Facebook and Instagram in Europe to comply with regulatory demands of the General Data Protection Regulation, privacy champion and long-time Facebook antagonist Max Schrems has engineered a legal challenge to the plan. Meta’s subscriptions are priced starting at 9.99 euros a month. Schrems responded to Meta’s announcement straightaway arguing that the fundamental right to privacy should not be contingent on financial resources, and that nyob, the privacy advocacy group he leads, would file a challenge. They have done so on behalf of a low-income Facebook user in Austria with that country’s data protection regulator. Along with the claim that avoiding online tracking for advertising should not be contingent on financial resources, they argue that Meta’s pricing far exceeds their reported per-user advertising revenues, and that if the model proliferates across all apps, it will be prohibitively expensive for most Europeans to avoid tracking. Meta responded that paid-versions of ad-free apps are a common online practice and their pricing squares with services like YouTube and Spotify.

Context – Fights over so-called “behavioral” advertising are almost religious in nature. For example, using data gathered online to direct ads is far more efficient and effective than “traditional” advertising. That simple fact means different things to different sides. Backers say it is key to small businesses and those with small advertising budgets having unprecedented success online. Many privacy advocates oppose effective advertising and see it as a sign of unwary online users being manipulated. Plus, there is the ongoing conundrum for targeted ad opponents that while everyone says they want advertising “privacy”, few actually choose to pay for it. Nyob says 97% of Europeans say they don’t want ad tracking, but that only 1% say they will pay for this privacy feature. We know European judges will weigh in. But nyob is also raising the issue of prices. This could be where EU Commission regulators step in as old-school price setters through their DSA and DMA roles.

Australia’s ACCC Says Digital Giants Threaten Competition and Consumer Harm

Report from Reuters

In Brief – The Australian Competition and Consumer Commission (ACCC) used its latest interim report in the Digital Platform Services Inquiry to raise concerns that the expansion of the wide-ranging digital ecosystems of Amazon, Apple, Google, Meta, and Microsoft can benefits consumers but also threaten competition and consumer harm. Mimicking “gatekeeper” language of regulators in Europe, the ACCC says the five US-based giants have established strongholds in sets of core services and attempt to grow into adjacent or complementary markets. While the report says, “Big is not automatically bad” and that this growth can provide benefits to consumers, it also sees a risk that this expansion may protect or extend market power through practices such as the bundling of products, pre-installation of apps and software, and default settings to limit customer choice or deter innovation at competitors. ACCC Chair Gina Cass-Gottlieb said, “Our proposed reforms include a call for targeted consumer protections and service-specific codes to prevent anti-competitive conduct by particular designated digital platforms.”

Context – Just a handful of years ago, with former Prime Minister Scott Morrison leading the government, and Rod Sims leading the ACCC, Australia was often at the forefront of efforts to aggressively regulate online platforms. Morrison rarely shied away from an opportunity to personally criticize tech giants. His government aggressively moved to mandate takedowns of abhorrent images following the 2019 Christchurch massacre, enacted an Online Safety Bill in 2021 regulating a broader range of objectionable content on social media, pressed aggressively on Big Tech antitrust, called for legislation to require social media companies to pierce online anonymity or face defamation risks for anonymous posts, as well as regulate how platforms deal with disinformation and misinformation. The digital policy that has resonated the most globally was the effort to compel Google and Facebook to pay Australia’s traditional media companies more money by enacting the News Media Bargaining Code. Admittedly, European governments have caught and past them on most digital regulatory measures.

Ask A Question!