When the Federal Trade Commission (FTC) issued a staff report in late October on the privacy and data collection practices of six large internet service providers (ISPs), the commission’s chair Lina Khan called the findings “striking.” Her own remarks are striking too.

After flagging structural issues in the marketplace, Khan asserted that “the Federal Communications Commission (FCC) has the clearest legal authority and expertise to fully oversee [ISPs],” and called for the FCC to reassert authority over ISPs “and once again put in place the nondiscrimination rules, privacy protections, and other basic requirements needed to create a healthier market.” Khan’s position was shared by FTC Commissioner Rebecca Kelly Slaughter, who wrote: “I hope the FCC is able to return ISPs to their proper classification as telecom services under Title II and to provide appropriate protections for these essential services.”

It’s not common for a regulator to disclaim authority. Regulatory humility can be a useful trait, but it is misplaced in this instance. For several reasons, it makes sense for the FTC to regulate ISPs with regard to privacy.

First, insofar as possible, a single set of privacy rules should apply to all sectors at the federal level, rather than add more patches to the crazy quilt of federal privacy regulation that exists in America today. It is confusing for consumers to have different privacy protections from businesses depending on their respective sectors—one law for financial services, another for communications companies, still another for health care, and no specific privacy law for most others. It is even more confusing for consumers if two directly competing services—such as location-based services offered by both “over the top” providers, like Hulu and Netflix, and ISPs—are subject to separate privacy regimes from multiple regulatory agencies. And it creates uncertainty or conflicting regulation for the ISPs themselves, many of which provide not only broadband internet access and other regulated communications services but also other services that are clearly subject to FTC jurisdiction, such as entertainment, content, advertising, and alarm services and other home automation.

A unified legal regime for privacy across sectors would be desirable, but very difficult to overlay onto existing regimes. This is why the report and draft legislative language I co-wrote with other Brookings scholars recommended that federal privacy legislation establish a federal commission to review existing laws and subsequently make recommendations to Congress “about how federal laws addressing privacy and data security may be harmonized.” In the meantime, it makes sense to avoid further fragmentation of federal privacy law.

Second, while the FCC and FTC’s differing legal authorities present a variety of issues discussed below, the FTC has much deeper and wider expertise when it comes to privacy. As Professor Chris Hoofnagle documents at length in his book Federal Trade Commission Privacy Law and Policy, the FTC has more than 100 years of experience in monitoring business practices and consumer protection, and has “evolved into the most important regulator of information privacy—and thus innovation policy—in the world” since the early 1990s. Hoofnagle joined leading privacy scholars Daniel Solove and Woody Hartzog in these pages in urging that, despite a short leash from Congress, ingrained timidity, and occasional capture, “the FTC is still the right agency to lead the US privacy regulatory effort.”  As a federal court of appeals put it in the 2018 FCC v. AT&T decision, “the FTC is the leading federal consumer protection agency and, for many decades, has been the chief federal agency on privacy policy and enforcement.”

Over the past two decades, the FTC has brought approximately 80 cases based on the handling of personal data against companies across a range of sizes and sectors: from Facebook and Google, to brick-and-mortar businesses, to small app providers and marketing companies. This caseload reflects that privacy has become a major focus of the FTC’s work in the 21st century digital economy. Just last week, the agency proposed a rulemaking process to impose stronger privacy and security protections for businesses, as well as prevent discrimination resulting from automated decision-making. While privacy has been a steady diet at the FTC, it has been only occasional fare at the FCC. The FCC has brought cases against common carriers and telephone marketers based on narrow laws that regulate telecommunications calling data, cable television subscriber records, and telephone marketing, though these recently resulted in a $200 million fine to Verizon, AT&T, T-Mobile, and Sprint for selling customers’ location information without proper security measures or consent. The FCC undertook broad rulemaking to apply privacy laws to ISPs following the adoption of the 2015 Open Internet Order, which subjected them to public service obligations applicable to communications common carriers under Title II of the 1934 Communications Act, as amended by the 1996 Telecommunications Act. But, Congress voted to override these privacy rules in March 2017, and so the FCC never carried out this broadened enforcement role.

Third, the views of Commissioners Khan and Slaughter do not reflect the significant progress toward comprehensive federal privacy legislation since 2016 and the prevailing thrust of the resulting legislative proposals to expand FTC authority. Federal bills or draft proposals—in particular, those from leadership on both sides of the aisle within the Commerce Committees in both houses of Congress that are the most likely vehicles for enactment of a federal law—predominantly place federal enforcement in the hands of the FTC. Senator Roger Wicker’s (R-MS) SAFE DATA Act would authorize $100 million for the FTC to exercise this authority. In addition, the $1.8 trillion economic reconciliation package, as passed by the House, includes a provision for $1 billion to establish and staff a new data protection bureau within the FTC.

Moreover, several privacy bills, including the SAFE DATA Act (but not Senator Maria Cantwell’s (D-WA) Consumer Online Privacy Rights Act), would override an exemption from the 1914 FTC Act to give the agency jurisdiction to enforce privacy violations by communications common carriers and non-profits. So long as that exemption is in place, there is some basis to Chair Khan’s contention that the FCC has clearer jurisdiction to monitor ISPs—if they are classified as communications common carriers, at least. However, that Title II classification remains uncertain. The congressional override of FCC broadband privacy rules, the subsequent repeal of the 2015 Open Internet Order by the Republican-majority FCC in 2017, the absence of a new net neutrality bill so far in the 117^th Congress, and delays in confirming a third Democratic commissioner to the current FCC could prevent or delay a new net neutrality initiative and FCC rulemaking.

Even if the current Congress, or a new Democratic majority at the FCC, pursues a new net neutrality order in due course, net neutrality is fundamentally about competition, not privacy; it is aimed at preventing ISPs from discriminating against competing service providers, such as online video service competitors like Netflix or YouTube. The FCC’s 2016 privacy regulations were an incidental effect of bringing ISPs under rules for common carriers, which was also aligned with the existing net neutrality provisions. The latter includes Section 222 of the Communications Act, which limits communications carriers’ use of “customer proprietary network information.” As a result, once the FCC deemed ISPs to be common carriers, the agency needed to address the application of Section 222 to ISPs, which protects the confidentiality of “customer proprietary network information” such as call records.

Congress can obviate this issue by giving the FTC jurisdiction over privacy for communications carriers—thus clarifying the FTC’s leading role on privacy. It is the path of least resistance toward enactment of comprehensive privacy legislation and a simpler, clearer way of protecting individual privacy.

Facebook, Google, T-Mobile, and Verizon are general, unrestricted donors to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the author and not influenced by any donation.

      

The troubling information provided by Facebook whistleblower Francis Haugen continues to generate new insights into the failure of the social media platform to police bad actors and moderate harmful content. As a result, the broader question of whether and how government intervention into platform governance should be pursued remains an ongoing topic of deliberation, with the latest in a series of Congressional hearings on the topic taking place last week.

However, in the U.S., these deliberations are unlikely to lead to action due to our strong First Amendment jurisprudence, which maintain extensive protections for both disinformation and hate speech. In addition, the unprecedented hostility that the Trump administration held for the news media has served as a powerful reminder of why we should be leery of any kind of new government intervention in the media sector. That being said, if, in the ultimate cost-benefit calculus, we see our commitment to First Amendment absolutism actually undermining the democracy that the Amendment is intended to protect, then perhaps some reconsideration of how we treat speech on social media may be warranted. One possible path forward in this reconsideration is to revisit how we regulate indecency in broadcasting.

Why indecency? And why broadcasting? Because broadcast indecency represents the only time in which the Federal Communications Commission (FCC) and the Supreme Court have agreed to the creation of a category of speech that exists exclusively within the confines of a specific medium from a regulatory and legal standpoint. Unlike, for example, obscenity, which is an unprotected category of speech regardless of how it is disseminated, indecency is only less protected within the context of broadcasting. There are no federal restrictions regarding indecency in any other communicative context. This is reflected in the FCC’s definition of indecency “material that, in context, depicts or describes sexual or excretory organs or activities in terms patently offensive as measured by contemporary community standards for the broadcast medium.” (emphasis added).

What relevance might a category of speech developed for an old and increasingly irrelevant medium have for the issue of social media regulation? From a substantive standpoint, not much. Indecency, as defined above, is at best on the periphery of concerns around social media, where issues of hate speech and disinformation are proving to be the most profound and impactful problem areas.

What is potentially relevant though—and perhaps worth considering—is the underlying notion that one or more distinctive categories of speech could be carved out exclusively for the social media context, in the same way that indecency is a regulatable category of speech exclusively within the broadcast context. Could we imagine a regulatory environment in which hate speech and disinformation continue to go largely unregulated except within the specific and narrow context of social media? This is a question I have been exploring as part of a larger research program that has been investigating whether regulatory frameworks and rationales developed within the traditional media sector might have useful lessons that can guide our approach to social media platforms.

In keeping with the model of media regulation that has been developed in the U.S., this kind of special treatment of speech within the exclusive context of a particular medium must be accompanied by a valid rationale. That is, what makes a medium sufficiently distinctive that it merits differential treatment from a regulatory standpoint? In the broadcast context, the two rationales that have been most relevant in relation to indecency are: 1) that broadcasters utilize a scarce public resource (the broadcast spectrum), and as public trustees of this resource, are thus entitled to lower levels of First Amendment protection; and 2) that broadcasting is “uniquely pervasive,” and this pervasiveness provides a rationale for treating broadcasting differently than other media. It is worth noting that the Supreme Court has rejected subsequent efforts by regulators to apply the indecency standard to telephony, cable TV, and the Internet.

I have argued at length elsewhere that a compelling case can be made that, like broadcasters, social media platforms utilize a public resource (in this case, aggregate user data) and so should be treated like public trustees—similar to how broadcasters must abide by certain conditions in exchange for their access to the collectively-owned broadcast spectrum. Whether social media platforms are “uniquely pervasive” in the ways that the FCC and the Supreme Court once considered broadcasting to be is another issue. The two media do share key criteria that the Supreme Court has brought to bear in characterizing broadcasting as uniquely pervasive. They are both free, widely available, easily accessible, and operate in such a way that unexpected, accidental exposure to harmful content is always a legitimate possibility.

Imagine for a moment that one of these rationales for treating social media differently from other media did gain traction. Might, then, allowing the government to play a more active oversight role over certain categories of speech in this narrow, but profoundly significant, context make sense? This more active role need not involve the kind of case-by-case adjudication and intervention associated with broadcast indecency, but rather, as those like Mark MacCarthy have advocated, some degree of accountability to a federal authority in regards to meeting independently-measured effectiveness thresholds for systems for policing disinformation and hate speech.

Certainly, the problems of hate speech and disinformation are not unique to social media. Hate speech and disinformation emanate from traditional and partisan news outlets, politicians, and political organizations, activists, and bad actors of various types. Yet social media represent an important mechanism by which the voices of all of these stakeholders are amplified, as platforms such as Facebook, Instagram, YouTube, and Twitter have evolved into some of the most pervasive, unconstrained, and engaging media outlets the world has ever known.

The bottom line is that a precedent has been established for constructing less-protected categories of speech exclusive to the confines of a particular medium. As we continue to struggle with the destabilizing repercussions of disinformation and hate speech, should policymakers and the courts consider creating explicit definitions of these speech categories that operate exclusively within the social media context? Could doing so provide a path forward for some form of intervention by which the government is better able to hold these platforms accountable for the content that they host and disseminate? This could certainly be considered an extreme, and ultimately misguided, approach to the problems of hate speech and disinformation, but in light of the magnitude of the challenges that we currently face, it may at least be a conversation worth having.

Meta, the parent company of Facebook and Instagram, and Google are general, unrestricted donors to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the author and not influenced by any donation.

      

On December 1, a federal judge in Texas issued a ruling blocking the state from enforcing its new social media law. Shortly after Texas Governor Greg Abbott signed House Bill 20 into law in September, NetChoice and the Computer and Communications Industry Association (CCIA) filed suit in federal court, arguing that it is unconstitutional.

Under HB 20, the largest U.S. social media companies “may not censor a user, a user’s expression, or a user’s ability to receive the expression of another person based on . . . the viewpoint of the user or another person.” This prohibition applies only to users who reside in, do business in, or share or receive expression in Texas.

In granting the plantiffs’ request for a preliminary injunction, Judge Robert Pitman of the U.S.  District Court for the Western District of Texas wrote that “HB 20’s prohibitions on ‘censorship’ and constraints on how social media platforms disseminate content violate the First Amendment.” Judge Pitman also noted multiple other First Amendment concerns, including what he characterized as HB 20’s “unduly burdensome disclosure requirements on social media platforms,” and the fact that HB 20 only applies to social media platforms with at least 50 million monthly active users in the United States. With respect to this size threshold, Judge Pitman wrote that:

“the record in this case confirms that the Legislature intended to target large social media platforms perceived as being biased against conservative views and the State’s disagreement with the social media platforms’ editorial discretion over their platforms. The evidence thus suggests that the State discriminated between social media platforms (or speakers) for reasons that do not stand up to scrutiny.”

So, what happens next? First of all, it’s important to note that a preliminary injunction is just that: preliminary. It does not mean that the plaintiffs have definitely prevailed in their challenge to HB 20. Rather, it indicates that the court concluded the plaintiffs have met the test explained by the Supreme Court in a 2008 decision: “A plaintiff seeking a preliminary injunction must establish that he is likely to succeed on the merits, that he is likely to suffer irreparable harm in the absence of preliminary relief, that the balance of equities tips in his favor, and that an injunction is in the public interest.”

Texas has already filed a notice of appeal, meaning that the decision to grant a preliminary injunction will be reviewed by the Fifth Circuit. The Texas case is following a similar trajectory to a case in Florida arising from that state’s enactment of a law targeting the largest social media companies. The same plaintiffs as in the Texas case, NetChoice and CCIA, challenged the Florida law, and achieved the same initial result: a preliminary injunction blocking its enforcement. That decision has been appealed by Florida and is currently before the Eleventh Circuit.

The coming months will thus see two different federal appeals courts weighing in on cases concerning one of the most important contemporary technology-related constitutional law questions: To what extent can the government regulate social media content moderation decisions without running afoul of the First Amendment?

While the specifics of the laws are different—the Florida law is aimed at preventing de-platforming of politicians, while the Texas law addresses content moderation more generally—they raise a set of overlapping questions about the limits of government power over the free speech rights of private entities. And while the current issue before the federal appeals courts is not the constitutionality of the laws themselves but rather the lower court decisions to preliminarily enjoin them, it is difficult to address the latter issue without considering, at least indirectly, the former. After all, each federal appeals court will need to evaluate whether a lower court was correct in concluding that the plaintiffs were likely to succeed in challenging a new state social media law on constitutionality grounds.

The Fifth and Eleventh circuits will likely do more than simply issue, without any substantive explanation, a simple thumbs up or thumbs down on the preliminary injunctions. Rather, in rendering their decisions, they may provide analysis that will shape future social media regulation attempts by state legislatures in the Fifth Circuit (which covers Louisiana, Mississippi, and Texas), the Eleventh Circuit (Alabama, Florida, and Georgia), and beyond. Even a circuit court decision upholding a preliminary injunction may provide guidance on the ways in which, at least in that circuit, a revised social media law might be more robust to attempts to enjoin it.

Empowered by this guidance, a state legislature could respond by crafting new legislation carefully designed to survive challenges to its constitutionality. In the long run, the most important legacy of the Texas and Florida social media laws may not be the laws themselves, but the way in which the jurisprudence they spur influences future legislative approaches to social media regulation.

Amazon, Apple, Dish, Facebook, Google, and Intel are members of the Computer and Communications Industry Association and general, unrestricted donors to the Brookings Institution. Amazon, Facebook, Google, and Verizon are members of NetChoice and general, unrestricted donors to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the author and not influenced by any donation.

      

Social media has become a hot topic over recent months, as the failure of companies to handle algorithmic discrimination, the spread of misinformation and the exploitation of children have translated to adverse societal harms. However, when considering how to regulate and resolve existing problems, it is interesting to examine how teens, the most tech-savvy population of all, have been navigating the internet. Which social media platforms do Gen-Zers prefer, and how do they navigate platform algorithms? How have online influencers played a part fighting back against vaccine misinformation and other societal injustices? What is the influencer industry, and in what ways are young influencers vulnerable to burnout and exploitation by social media companies and talent agencies?

To answer these questions, Samantha Lai will be joined by Taylor Lorenz, technology reporter for the New York Times and affiliate at the Harvard’s Berkman Klein Center for Internet & Society, to discuss how young people use the internet and why it matters.

You can listen to the episode and subscribe to the TechTank podcast on Apple, Spotify, or Acast.

TechTank is a biweekly podcast from The Brookings Institution exploring the most consequential technology issues of our time. From artificial intelligence and racial bias in algorithms, to Big Tech, the future of work, and the digital divide, TechTank takes abstract ideas and makes them accessible. Moderators Dr. Nicol Turner Lee and Darrell West speak with leading technology experts and policymakers to share new data, ideas, and policy solutions to address the challenges of our new digital world.

      

The disclosures from the Facebook Papers have led to a flurry of legislative proposals on Capitol Hill to address data use, kids’ online safety, and malicious content. The single most effective step Congress can take is to enact comprehensive privacy legislation to address the explosion of digital information not covered by existing, narrower privacy laws.

Congress should not let this latest “Facebook moment” pass without meaningful action. Every day that passes without a baseline privacy protection law in effect is another day that not just Facebook, but a multitude of businesses, collect and use data generated from billions of devices. IBM estimates that the world generates 2.5 quintillion bytes of data per day – that’s 2.5 followed by 18 zeroes.

This volume will soon explode even further as data-intensive technologies power augmented and virtual reality environments (the “metaverse” to which Facebook’s new corporate name refers), and autonomous vehicles in the streets communicate with sensors that monitor traffic and pedestrian safety.

It is past time for the United States to require businesses to use data responsibly and give individuals rights in data about them, safeguard that have been absent and that would protect everyone in America.  Every major democracy in the world has passed legislation to set boundaries on business data collection – and this month a new commercial privacy law took effect in China.

After the Cambridge Analytica stories in 2018, several congressional committees hauled in Facebook founder Mark Zuckerberg for back-to-back hearings carried live on cable news. Since then, Facebook has been a subject of some 70 hearings on privacy, competition, content moderation, misinformation, security, and diversity and inclusion, with 17 Facebook-affiliated witnesses.

Galvanized by Facebook’s sharing of data with Cambridge Analytica, Congress made a promising start on comprehensive privacy legislation. Leading members called for legislation, several introduced bills, and multiple committees and working groups got to work toward bipartisan legislation.

When it came to the hard bargaining, however, the early promise waned. Even though flagship bills from the leaders of the Senate Commerce Committee had encouraging areas of overlap, bipartisan agreement stalled. Ideological red lines that have frozen progress include whether a federal law should preempt state privacy laws in part or altogether, and whether to allow lawsuits by individuals for violations of privacy rights.

There is ample room for compromise on these issues, but little sense of urgency to do so. As privacy work slowed, Congress’ attention has turned to the many other concerns about Facebook and tech platforms, including their impact on children, free speech and disinformation, competition, and jobs.

Now the Facebook files have rekindled the privacy debate. The Senate Commerce Committee held its first privacy hearing this year in September. Committee Chair Maria Cantwell (D-WA) and Ranking Member Roger Wicker (R-MS) both affirmed their interest in working together on comprehensive legislation. At this hearing, Senators and business witnesses publicly softened their differences on the fractious issue of private lawsuits.

The two of us have been deeply involved in the public discussion of privacy legislation for more than a decade. We speak often with congressional staffers and members as well as advocates for industry, consumers, privacy, and civil rights. We have never seen passage of a broad privacy law as close as it is now, and no other issue targeting the major tech platforms is anywhere near as ripe for action.

The work already done in Congress and through laws passed in California, Colorado, and Virginia have laid a foundation for passage of federal legislation. Both -industry and advocates for privacy, consumer protection, and civil rights are ready to compromise on key issues when and if Congress moves toward passage of a privacy bill.

It will take more than privacy legislation to solve the wide-ranging challenges that social media has created and amplified. But protecting people’s personal information is the place to start. Congress should finish the job of enacting a privacy law that began with its first Facebook hearings in 2018.

Facebook and IBM are general, unrestricted donors to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the authors and not influenced by any donation.

      

In October, the Forum for Cooperation on Artificial Intelligence (FCAI), a multistakeholder dialogue among high-level government officials and experts from industry, civil society, and academia, released an interim report taking stock of the current landscape for international cooperation on AI and offering recommendations to make further progress.

FCAI publicly launched the report as part of Brookings’ Global Forum on Democracy and Technology event, Aligning technology governance with democratic values. UK Secretary of State Digital, Culture, Media and Sport, Nadine Dorries, praised the “excellent” report as a “helpful step in [the] process” of building international AI collaboration while discussing her government’s role in its presidency of the G7 group and its upcoming Future Tech Forum. To discuss the report, Brookings co-authors Cam Kerry and Josh Meltzer, and Andrea Renda of the Centre for European Policy Studies (CEPS) welcomed a panel featuring representatives from the governments of Australia, Canada, and the United States, as well as industry representatives from IBM and Twitter.

While the entire event and panel discussion around the report can be found here, for some unfamiliar with the FCAI, this blog will serve as an introduction to the Forum and the new report. Specifically, it will provide background on the creation of the FCAI and preview key elements of the report, including the arguments for international cooperation on AI, the current international AI policy landscape, and the list of proposed recommendations with which the FCAI intends to shape future dialogues on the issue.

What is the FCAI?

As the strategic, economic, and social significance of artificial intelligence has become widely recognized in recent years, governments, industry, and other international stakeholders have started to develop individual strategies to capitalize on opportunities and address challenges.

Since 2017, when Canada became the first country to adopt a national AI strategy, at least 60 countries have adopted policies in some form—declarations, frameworks, industry guidance, or principles—focused on artificial intelligence. Industry leaders in the tech sector have taken similar steps to codify their approaches to AI, working toward responsible, trustworthy, and ethical use and outcomes.

As these efforts across stakeholders became increasingly global—and their outputs increasingly robust – Brookings and the Center for European Policy Studies (CEPS) worked together on a deeper exploration of future harmonization mechanisms and established the FCAI in early 2020.

Over the past 18 months, the FCAI has held nine AI Dialogues, bringing together hundreds of participants for high-level, multistakeholder roundtables featuring government officials from the UK, U.S., EU, Canada, the UK, Japan, Australia, and Singapore, along with leading experts from academia, the private sector, and civil society. The FCAI report, Strengthening international cooperation on AI: A Progress Report, summarized below, is the culmination of the first eight of the nine roundtables.

Why international cooperation on AI is important

The report makes a strong case for international cooperation on AI. Grounded in the concrete realities of AI development, international trade, and democratic values, Section 1 of the text delineates both the negative consequences of an international landscape lacking in cooperation and the benefits of increased partnership. It also highlights potential positive impacts of AI to address global challenges like climate change or pandemic preparedness. The report argues that “no country can go it alone,” and demonstrates how powerfully a collaborative international framework could influence the positive trajectory of AI development and deployment. The necessity and benefits of a collaborative approach are exemplified by the increasingly global AI research-and-development landscape, proposing that cooperation across international teams has the potential to facilitate resource-intensive research and enable developers to upscale projects.

The international state-of-play

Section 2 of the report provides an overview of the state-of-play in countries currently participating in FCAI, detailing both domestic outputs such as AI strategies, industry guidance, and investment data, as well as concrete commitments to engagement on the international level. This section also charts the broader international terrain, demonstrating the wide range of international bodies (the 17-nation Global Partnership on AI and OECD, UNESCO, WTO, APEC), international standards organizations (ISO/IEC, IEEE, ITU-T), and non-governmental bodies (NYU’s AI Now Institute, the World Economic Forum, the private sector Partnership on AI) that contribute to the international discussion on AI.

Recommendations

Finally, the FCAI report presents fifteen specific recommendations for further developing international cooperation on AI. These recommendations fall into three broad categories: regulatory alignment, research and technology-driven standards development, and joint research and development. These broader categories are discussed below, highlighting key recommendations that undergird others and showcase the larger intent of each grouping.

The first ten recommendations of the report discuss improvements for regulatory alignment, and approach cooperation on AI as an incremental process where lighter forms of cooperation can compound over time and lead to more comprehensive practices. As a foundation, Recommendation 1 calls on governments to embed their commitments to cooperation on AI into their domestic strategies. Other recommendations in this category continue to lay groundwork for efficient communication and collaboration, such as agreeing on a common definition for AI, further aligning domestic frameworks of ethical principles, and establishing redlines in AI development to preserve democratic values and protect individual rights.

Recommendations 11 through 14 focus on developing the capacity for cooperation on AI standards in international standards bodies, such as ISO/IEC, IEEE, and ITU-T. Similar to the incremental approach proposed by the recommendations on regulatory alignment, recommendation 11 advocates for a “stepwise” approach that begins with foundational standards around definitions and terminology that can be applied in a horizontal, cross-cutting fashion, establishing a firm foundation on which new standards can be built as technologies mature.

The unique challenge of China, which is discussed throughout the report as both a foil in its techno-authoritarian approach to AI and as an inescapable partner for international collaboration, is also addressed concretely in the standards recommendations. Warning that international standards bodies should not become a proxy frontier for geopolitical competition, FCAI’s recommendation 13 calls for government participating in the forum to coordinate on international standards development in a way that encourages Chinese participation, but safeguards the industry-led, research-driven approach used by standards bodies. This approach prioritizes technical knowledge over political posturing.

In the final category, the single recommendation 15 calls for the development of common criteria and governance arrangements to facilitate international collaboration on large scale R&D projects. In addition to functioning as a vehicle for streamlining AI cooperation on challenges of global scale and significance, working on pressing issues where the outputs are public goods can operate as a high-incentive sandbox that allows governments and other stakeholders to find common ground while working in a collective environment.

Moving forward, FCAI intends to continue hosting dialogues, using this report and the recommendations within to delve deeper into ongoing conversations and approach new topics with a stronger foundation.

IBM is a general, unrestricted donor to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the author and not influenced by any donation.

      

The COVID-19 pandemic has facilitated the unprecedented growth of telehealth, with a 78-times increase in use from February 2020 to April 2021. And even as widespread vaccination of the American public has lowered the risk for people to go back to in-person doctor’s appointments, many continue to utilize telehealth services. With telehealth here to stay, what can and will the Biden administration do to not only harness its power, but also advance digital equity to create more efficient and effective use by health practitioners and patients?

On this episode of TechTank, Brookings researcher and guest host, Samantha Lai, is joined by Nicol Turner Lee and Niam Yaraghi from the Brookings Institution to talk about their upcoming paper on telehealth and the important elements of a telehealth 2.0 roadmap coming out of the pandemic. As policymakers, advocates, industry professionals, and patients consider the next iteration of telehealth, how will past legislation that limited the use and reimbursement of telehealth be treated? What have been lessons during the pandemic that can be applied to both the provision and quality of telehealth and other forms of remote care? Will care modality matter in the next iteration of telehealth, especially if it means more patients – regardless of proximity to medical institutions or doctors – can be treated equitably? How will other elements of the health care system, like Health Information Exchanges, be modernized? And will telehealth be successful if the U.S. does not close the digital divide? All of these are important questions to answer in order for the nation to better chart a path to the responsible, inclusive, and flexible use of telehealth going forward.

You can listen to the episode and subscribe to the TechTank podcast on Apple, Spotify, or Acast.

TechTank is a biweekly podcast from The Brookings Institution exploring the most consequential technology issues of our time. From artificial intelligence and racial bias in algorithms, to Big Tech, the future of work, and the digital divide, TechTank takes abstract ideas and makes them accessible. Moderators Dr. Nicol Turner Lee and Darrell West speak with leading technology experts and policymakers to share new data, ideas, and policy solutions to address the challenges of our new digital world.

      

When digital mobile phone technology was first introduced in the US, electric wheelchairs began behaving erratically. The pulsing signal interfered with their controls. The solution: simple shielding to stop the interference.

When phones using the international GSM digital standard were first introduced in the US, hearing aids would buzz. The hearing aids, which had been designed for the analog world, were suddenly confronted by a new digital reality. The solution was once again updating the old way of doing things to recognize the new environment.

And if you want to talk life-and-death, how about pacemakers? Again, early in the digital phone era these life-savers could malfunction when hit by a cell phone signal. The short-term solution was for doctors to tell pacemaker patients not to carry their phone in the shirt pocket. Long term, shielding solved the problem.

These stories all come to mind as the Federal Aviation Administration (FAA) has objected to the Federal Communications Commission’s (FCC) authorization to use newly opened airwaves for 5G networks. Their concern is that the 5G signals could possibly interfere with the radio altimeters used in automated aircraft landings.

The 5G-aviation safety issue combines the two most important components of public policy decision-making: public safety and national security. No one can question the importance of the safety of passengers on commercial and private aircraft. Similarly, everyone recognizes that the United States is in a technological horse race with China, which has been able to reap the rewards of its embrace of 5G networks.

The Shared Spectrum Resource

The forward march of technology has once again tripped over the old way of doing things.

The airwaves (often referred to as “spectrum”) are a shared national resource that is subject to the pulls and tugs of changing technologies.

As these technologies change, the assumptions that previously governed the spectrum-based environment also change. What was an adequate product design in the earlier era—like the wheelchairs, pacemakers, and hearing aids—may require a redesign. Neither the device manufacturer nor the spectrum users were at fault. The device manufacturers built to the realities that existed when the product was designed. The new spectrum users were building into a new world without any intention of causing harm.

I was president of the wireless industry trade association CTIA during each of the earlier interference issues. I remember the surprise when the issues first developed. I recall the headlines and the emotion. I remember how some companies tried to profit from the problem.

Years later, I was Chairman of the FCC, and know the agency’s responsibility to protect consumers while keeping the United States at the forefront in wireless applications and services. I also recall how the FCC was constantly being called upon to play referee between various users of spectrum, and how easy it was to allege “interference” as a competitive strategy or a financial tactic.

From both those experiences, I learned that the challenges created by technological change can be solved if people of good will leave their respective corners and PR campaigns, and instead come together in a shared commitment to a solution. Beyond the issue of goodwill, however, is the need for federal leadership. Such leadership in the form of a national spectrum policy was notably absent during the Trump administration.

A Lack of Leadership

The airwaves are a shared national resource and exist in an environment of continual technological change and marketplace development. As a result, the federal government must have a set of underlying policies to guide the difficult and highly technical decisions about spectrum allocation and its ultimate effects on incumbent as well as new users. Unfortunately, the Trump administration had no such unified spectrum policy; as a result, policy ended up being made by individual agencies.

Less than 14 months after President Obama took office, his administration produced an integrated spectrum plan and broadband plan. It was not until 20 months into the Trump administration that there was even an attempt to create a national plan for spectrum usage. In October 2018, President Trump ordered the plan to be available in six months. But it never happened.

The Department of Commerce’s National Telecommunications and Information Administration (NTIA) is supposed to be the telecommunications advisor to the president. It was NTIA that was tasked with developing the national spectrum plan that never was. Unfortunately, and reportedly as a consequence of a spectrum dispute, the NTIA head was axed and the agency remained without a permanent leader for the last 20 months of the Trump administration.

The consequence of this absence in both framework and leadership meant there was no underlying rationale nor consistent team to adjudicate among the various spectrum claimants. This left government agencies free to advocate their own spectrum policies. In such a situation, it is only natural that the individual agencies would retreat into their comfort zones and view spectrum only within their parochial interests. No one ever wants to change the way things have always been done by revisiting the best use of spectrum. In the absence of White House leadership on a national policy, agencies such as the FAA, Department of Defense, and FCC quite naturally prioritized the interests of their own constituencies.

Avionics and 5G

The spectrum used by aeronautical navigation systems as well as so-called C-band wireless are internationally allocated. On the spectrum allocation chart, the aeronautical frequency allocation runs between 4.2 and 4.4 gigahertz (GHz). One of the key uses of the aeronautical allocation is the transmission of information to and from aircraft altimeters, especially when they operate below 2500 feet, to facilitate computer-assisted landings. Next to that allocation is the C-band spectrum used for 5G. In the U.S., C-Band use is authorized for between 3.7 and 3.98 GHz.

The airlines and associated industries are warning that 5G networks operating in the C-band “have the potential to cause harmful interference to radio altimeters.” Their concern is that the radios being used with the altimeter may not appropriately filter out signals lapping over from another part of the spectrum (called spurious emissions). In response, the FAA issued a Special Airworthiness Bulletin (SAIB) to airlines and pilots to “be prepared for the possibility that interference from 5G transmitters and other technology could cause certain safety equipment to malfunction.” The Canadian government responded by restricting C-band usage around airports.

Two words are central to the statements of the airlines and FAA. The airlines talk about the “potential” of interference. The FAA talks about the “possibility” of interference. Clearly, the safety of air traffic requires mitigating even the “potential” or “possibility” of problems. Yet clear heads are needed to separate what is only hypothetical possibility based on worst-case assumptions from what is highly probable based on real-world use.

This is where the spectrum management expertise of the FCC is essential. The FCC is the nation’s spectrum referee on commercial spectrum interference matters. As such, the agency’s engineers are used to constantly dealing with the evolution from one generation of spectrum-related technology to another. It is a challenge made all the more difficult by the need to balance the national interest in technological advancement with the private interest of manufacturers and users of equipment designed for another spectrum environment.

What Happened with 5G and Altimeters?

The 5G technology is built to exacting standards. There are no official altimeter standards. As the FCC was working on reallocating portions of the C-band to delivering 5G signals, the issue of potential interference to radio altimeters became a matter of great discussion. But there were no common altimeter standards to measure against.

The aviation industry submitted a study by the Aerospace Vehicle Systems Institute (AVSI) that simulated worst-case 5G signal emission and its impact on avionics. The FCC’s analysis of the AVSI study found, “there may be a large variation in radio altimeter receiver performance between different manufacturers.” In other words, some altimeters were equipped with radio receivers with good filters to protect against spurious emissions, while others allowed signals from outside the 4.2 to 4.4 GHz allocation to intrude.

A 5G proponent, T-Mobile, submitted a study by Alion, an engineering firm critiquing the AVSI study. This analysis found the assumptions used in the study to be extreme, thus leading to extreme conclusions. In addition, the Alion analysis concluded, two of the test altimeters had failed “due to interference from other altimeters,” not 5G interference. After reviewing this study, AVSI told the FCC “further analysis is required to consider more sophisticated propagation models.”

The FCC’s Report and Order making the C-band available for 5G directly addressed this issue. The FCC engineers concluded the “AVSI study does not demonstrate that harmful interference would likely result under reasonable scenarios” or even “reasonably ‘foreseeable’ scenarios.”  The FCC encouraged the aviation industry “to take account of the RF [radio frequency] environment that is evolving.” In other words—like wheelchairs, pacemakers, and hearing aids— to recognize that what was “good enough” design in a previous spectrum environment could be affected by the new environment.

Nevertheless, the FCC created a guard band between the 5G spectrum and the avionics spectrum in which 5G was forbidden. Boeing, in a filing with the FCC, had proposed just such a solution. The Boeing proposal was to prohibit 5G “within the 4.1-4.2 GHz portion of the band.” The FCC agreed and then doubled the size of Boeing’s proposed guard band to a 220 MHz interference buffer between the upper 5G usage at 3.98 GHz, and avionics usage at 4.2 GHz.

When the FAA issued its bulletin, the 5G industry pulled back the C-band launch. Originally planned for December 5, the launch of service in the new spectrum was postponed by 30 days.

The aviation industry, in a letter to the White House National Economic Council (NEC), has pledged to work diligently “to develop new standards, equipment, and aircraft/helicopter integration solutions.” Interestingly, according to a Reuters report, the White House “reviewed the FAA safety bulletin before it was cleared for release.”

Next Steps

The resolution of this issue should not be a drawn-out process. The Biden White House is now involved and should be the driving force. Acting quickly to convene the parties is an important step to keep them from retreating into their own corners. The White House should take the aviation industry up on its pledge to “develop new standards.”

The working group should have a tight deadline to report its conclusions and be true to its name: a working group, not a study group, nor a debating society. The physics involved in this situation are well known. The mitigation techniques are well known. The standard-setting process is well known. The importance of getting 5G up and running while protecting flyers is well known.

The Biden administration has prided itself on being science-based. The science here is pretty clear—it is hard to repeal the laws of physics. The real politick of this comes down to the costs of fixing the altimeters, just like the wheelchairs, hearing aids, and pacemakers were fixed. As the FCC engineers concluded, “well-designed equipment should not ordinarily receive any significant interference (let alone harmful interference).”

Let’s hope this is more than a gambit to hold 5G hostage to get someone to pay to fix the problem altimeters. There are only three sources of such funds for the aviation industry. The government could pay out of the almost $82 billion generated by the sale of licenses to use the C-band; that would probably require an act of Congress. The wireless industry could pay an additional tariff on top of the billions already spent for spectrum the government said would be ready for use on December 5. The aviation industry, having known for some time of the new 5G allocation, could pay to fix the offending altimeters.

More generally, this brouhaha also highlights one more area where the Biden administration needs to repair what was left behind by the Trump administration: the lack of a spectrum plan for the nation. The 21^st century will be the wireless century. Already work is underway on 6G. The use of the spectrum necessary for the connecting pathway of the new era requires a going in strategy rather than a policy that distributes policymaking on an ad hoc basis.

T-Mobile is a general, unrestricted donor to the Brookings Institution. The findings, interpretations, and conclusions posted in this piece are solely those of the author and not influenced by any donation.

      

As digital technology accelerates, there are questions about who is most likely to lose jobs due to automation and what the overall future of the US economy looks like. These questions are worth asking—particularly after a pandemic that appears to have hastened the automation of many tasks in American industries. Yet research on automation has so far centered almost entirely on the presence of digitalization, automation-potential estimates, the relationship between technological change and macroeconomic conditions, and tech’s impact on inequality and wage divergence.

Since technological change often is quite disruptive and spurs economic and political shocks, it is vital for researchers to study the attitudes of the individuals most vulnerable to new technological shifts. In doing so, researchers can gain a humanizing window into how these shifts—so crucial to capitalism’s advancement—are borne out in the beliefs, characteristics, and fears of the individuals most likely to experience disruption due to automation

The results of my work offer an ominous window into how technological change may correspond with despair, radicalization, and democratic erosion. This study shows that the Americans whose occupations have the highest automation potential tend to have a dark and cynical view of politics, the economy, the media, and humanity. They comprise a traditional working class that is politically left-leaning on economic issues and slightly right-leaning on socio-cultural ones. Although they are moderately more likely to support Democrats, they are increasingly likely to support Republicans since 2000. These often-economically-vulnerable Americans are deeply pessimistic about the state of the world and politics; they also have a tendency to vote against their economic interests and become more authoritarian in their outlook.

Technological Change Often Causes More Inequality

My work examines studies on the relationship between economic upheaval and radicalization, polarization, and revolution. In addition, I look at how technological change impacts inequality, wage divergence, and job polarization. Works from David Autor, Robert Allen, and Daron Acemoglu, in particular, have been able to illustrate the ways that digitalization can catalyze labor substitution, the erosion of the middle class, and greater income inequality.

It seems clear that technological advancement can both stifle real wage growth for many workers while simultaneously increasing returns to capital by making labor more productive. This tends to benefit the owners of productive capital and some workers whose skills are complemented by the new technology. Moreover, when productivity tends to rise faster than wages, inequality increases almost by definition since new GDP gains accrue primarily amongst capital owners, rather than workers. Technological change can additionally lead to a hollowing out of middle-wage work, in turn producing a smaller middle class. Digitalization appears to be inducing dis-equalizing shifts through all of these mechanisms.

Americans Susceptible to Automation Have a Dim View of the World

My study merges American National Election Survey data from 1990-2016 alongside the automation-potential estimates that are produced by McKinsey and presented by Mark Muro in a 2019 report. Due to the considerable size of this dataset, I utilized 11 indexed variables to proxy for specific general characteristics (see Figure 1). I then looked at the relationship between the automation potential of an occupation and the beliefs of workers—for example, if a person believes the minimum wage should be increased or lowered.

First, I look at the demographic characteristics of the Americans most susceptible to automation.

Table 1: Demographic Summary Statistics

Table 1 clearly demonstrates that Hispanic individuals—followed by Black and Asian individuals—are more likely to work in jobs with the highest automation potential; women, too, are slightly more likely to fall into the high automation pool. This corroborates Mark Muro’s 2019 findings. Beyond this, the automation potential of the average worker tends to be higher in the South, among the lowest third of the income distribution, and generally among people with less education.

Yet what are the beliefs of these automation-susceptible individuals? In Graph 1, I look specifically at this question by offering a brief portrait of the views that the Americans most susceptible to automation are likely to hold.

As the model shows, Americans who work in an occupation that has a high susceptibility to automation tend to be less politically engaged, less supportive of the media, and more pessimistic about the nature of politics and the power of their voice. Although highly automation-susceptible individuals show a slight preference for Democrats, this preference is waning; in the years since 2000, it has fallen below the average support for Democrats amongst the broader US voter population. This indicates a possible shift away from the Democratic Party in the last two decades, as both political apathy and engagement have worsened.

Moreover, Americans with a high susceptibility to automation since 2000 also are nearly 5 percentage points more likely to express racist, antisemitic, homophobic, or sexist attitudes. This increase in intolerance is even more pronounced when isolating for more specific categories of racial animus; although many of the same individuals who hold racist views of Hispanics will also hold racist views of Black Americans, for example, this is not necessarily always true. The 5 percent figure thus underrepresents the relative increase in intolerance amongst automation-susceptible individuals.

Beyond this, Americans whose jobs are highly susceptible to automation are less supportive of globalization and immigration; however, they hold more left-wing economic views on average than the median working American. Automation-susceptible individuals also tend to be slightly more culturally conservative since the 2000s. Finally, highly automation-susceptible individuals are more likely to have a cynical view of human nature (e.g., believing people are fundamentally dishonest) and to express authoritarian tendencies—valuing obedience, for instance, over individuality and expression. In Table 2, I offer a breakdown of these results between different demographics of Americans.

Table 2: Difference in Means of High & Low Automation Groups

Some key findings here are that white Americans with a high automation potential diverge from other racial groups in their slight preference for Republican candidates and conservatives over Democratic candidates and liberals. White Americans whose jobs are highly susceptible to automation are also more likely to have a grim view of their lives and to be opponents of immigration and international trade than whites who are not as vulnerable to automation. Highly automation susceptible individuals of all demographics, however, are more likely to hold authoritarian views than low automation susceptibility groups.

The Evolving Worldview of America’s Working Class

My results point to three important trends in American politics and the economy. The first is that automation-susceptible individuals are indeed a demographic category with a sufficiently unique set of beliefs and characteristics. Second, individuals highly susceptible to automation are more likely to hold progressive economic beliefs, in line with their interests. Finally, highly susceptible individuals are becoming more despairing and more likely to vote for Republicans and against a progressive economic agenda.

One possible explanation for this is that the demographic most likely to be harmed by technological-change-induced labor shocks are increasingly voting based on cultural, rather than economic, issues. If this thesis is true, it would corroborate the notion that a distinct portion of the American working class has shifted away from the Democratic Party as the cultural wars came to dominate public debates. My results thus suggest that a combination of progressive economic and conservative cultural policies may prove most attractive to Americans whose jobs are highly susceptible to automation.

      

Methane’s march from obscurity to center stage at the Glasgow climate summit reflects its growing recognition as an intensive global warming driver. President Joe Biden launched a Global Methane Pledge shortly before COP 26, securing dozens of national commitments to reduce global methane 30 percent from 2020 levels by 2030.

Methane is not, however, a new climate concern or a new arena for multi-national policy. Five years ago, American, Canadian, and Mexican leaders embraced new continental pledges to address climate change. A North American Leaders Summit included vows to reduce methane releases from oil and gas production by 40-to-45 percent from 2012 levels by 2025.

Part of the enthusiasm surrounding this tri-national commitment stemmed from a recognition that energy-sector methane ranks among the technologically easiest and lowest-cost greenhouse gases to mitigate. Technologies to measure and monitor methane and capture it for use as natural gas are increasingly available, maturing rapidly, and often highly cost-effective. Captured methane retains considerable commercial value if used as natural gas rather than wasted.

Given these complementary factors, methane is routinely characterized as “low-hanging fruit” for climate policy. Perhaps this contributed to the big grins that Justin Trudeau, Enrique Peña Nieto, and Barack Obama displayed toward the end of their “Three Amigos Summit.” Five years later, Biden will host his own North American summit, joining Trudeau and Mexico’s Andrés Manuel López Obrador. Some reference to climate and energy is likely. But what happened to previous summit pledges?

North America still lacks a credible measurement system for methane releases. An ever-growing body of research using advanced technology routinely finds consistent downward bias in official data published in all three nations. Progress is being made by groups using satellites and related technology to monitor portions of the methane supply chain, but government-reported methane numbers remain suspect.

Policy development to mitigate methane releases has proven highly uneven. In the United States, Obama-era efforts to apply the Clean Air Act to methane collapsed after a withering counter-offensive by industry, states, and the Trump Administration. With a few notable exceptions, most notably Colorado, production states have subsequently stood pat on methane, reflecting the historic reluctance to upset production firms. Industry performance on employing best practices for technology and emission minimization continues to vary dramatically as best we know.

Since Biden took office, Congress has restored Obama-era provisions for new drilling operations. A methane fee remains under Build Back Better Act consideration, albeit weakened from its initial form and linked to funding to subsidize industry acquisition of mitigation technology. New federal funding is also in the offing to remediate abandoned and orphan wells, given past and present industry and state aversion to rigorous bonding provisions. The Environmental Protection Agency has just released proposed rules for expanded methane oversight, but these now begin a perilous political and legal path forward. In short, the United States has begun revisiting methane policy but with a very uncertain future after a squandered half-decade.

Mexico initially appeared poised for methane leadership, including through widely acclaimed 2018 legislation. But implementation has advanced glacially and López Obrador’s 2018 election triggered a profound Mexican energy policy pivot. Earlier support for rapid renewable energy expansion has been supplanted by all-out oil pursuit that restores Petróleos Mexicanos (Pemex) as the preeminent national energy force. Mexico holds an unenviable record of producing significant declines in oil output alongside major methane release increases. Expanded methane waste has fostered growing reliance on imported American natural gas.

The 2016 summit, however, was not a complete flop. Canada largely honored its part of the bargain while its two continental partners shirked. Trudeau prioritized methane regulation, leading to 2018 federal standards. This included tri-annual leak detection and repair, flaring and venting restrictions, and transition to low-emission pneumatic technology at production sites. The government has pledged to go farther in coming years.

Canada has framed methane policy as a complement to other climate initiatives such as its robust carbon pricing program in preparing Glasgow pledges. Its federal government lacks authority to impose national policies and so negotiated “equivalency” agreements with major production provinces. British Columbia promptly submitted an equivalency proposal, leading to early 2020 approval. According to a Pembina Institute analysis, “BC regulations have strong elements, representing best practices and in some cases exceeding the federal rules.” British Columbia also adopted legislation to address its growing orphan well backlog, blending expanded regulations with a new levy on production to assure funding. Alberta and Saskatchewan similarly produced equivalency plans but these were more modest than British Columbia’s, resulting in longer negotiations and some variation from federal intent. Ultimately, new agreements have been approved, lubricated by expanded federal funding for industry loans and grants to ease technology transition costs.

Canada emerges a half-decade after the 2016 summit with an operational federal plan that appears to be reducing emissions despite increased energy production. There remain uncertainties as to whether Canada can honor 2025 reduction pledges, given provincial inconsistencies and enduring doubts over release data accuracy. Nonetheless, while all three nations have embraced the Global Methane Pledge, only Canada has a clear forward path to address the energy component of this initiative.

Methane will not dominate the coming continental deliberations, given the many issues that will compete for summit attention. More broadly, the way in which other places such as China, Europe, and Russia deal with methane remains unclear. The new global pledge also includes other sectors such as agriculture, which poses different technology challenges and represents a difficult constituency for political leaders to address.

But one foundational step would entail North American commitment to employ state-of-the-art technology to systematically measure and report methane emissions. Both government and business can build on the established continental model for toxic emission releases and expand its application. As long as methane continues to lack clear measurement and oversight, multi-national pledges will mean little to long-term environmental sustainability.  North America could become a model for global efforts to deliver on the new international pledge.