The Narrative is a twice-monthly survey of key developments in Internet governance. 

This time we take up Facebook’s struggles with content moderation, the geopolitics of 5G, and coordinated efforts by governments to limit encryption.

Platforms and Content Moderation 

Platforms continue to struggle with the controversies and contradictions of content moderation, even as numerous legal and policy initiatives attacking Section 230 protection are underway, such as a new round of Congressional hearings October 28. Facebook chief executive Mark Zuckerberg has signalled his support for Congress updating Section 230 “to make sure it’s working as intended.” 

Facebook’s new Oversight Board, hyped as a “Supreme Court,” has announced that it is accepting appeals from individuals objecting to the removal of content. Initially the Board’s authority is limited to reviewing content removals, and it cannot take up appeals from users whose accounts have been suspended or cases where content that is in violation of Community Guidelines has not been removed. The Board has given itself a 90-day window to start reviewing cases, which means it will not be taking up cases related to the upcoming U.S. election. Although Facebook can refer cases to the Board for an expedited review both on an ongoing basis and under exceptional circumstances, Brent Harris, Facebook’s director of governance and global affairs, has clarified that the company would not submit a case for expedited review before the November 3 vote.

The conditions and processes for Facebook referrals and expedited hearings remain unclear. For example, the Board should be addressing controversies like the decision of Facebook’s public policy chief for India to exempt posts by members of the ruling BJP party and Hindu nationalists from takedowns for violating the company’s policies against hate speech or dangerous organizations and individuals. Although the India policy head in question has since stepped down, uneven enforcement of content policies continues to degrade trust in platforms. While the Oversight Board’s bylaws state that content that Facebook has ultimately allowed to remain on the platform will be reviewable “in the future”, the India hate-speech case would not be heard by the Board unless Facebook itself brings it forward. As things stand, the company has no incentives to appeal content moderation decisions of employees and teams, particularly when those decisions protect the company’s business prospects. 

These are early days and it remains to be seen what kind of accountability the Board will bring to Facebook’s content moderation, which has been lacking in both human oversight and options for appeals. The Wall Street Journal reported more than 150 pieces of content that Facebook later confirmed violated its rules, and found that the company’s review system allowed the material—some depicting or praising grisly violence—to stand more than three-quarters of the time. Facebook’s August 2020 transparency report shows there were zero appeals on Instagram during the second quarter of 2020 and very few on Facebook. Due to the pandemic the company has ended up curtailing the process for users to appeal against removal of content and get the content reinstated following review by a human moderator.

A few other stories of note. Facebook is attempting to shut down a New York University Ad Observatory that has recruited more than 6,500 volunteers to collect data about the ads Facebook shows them by using a specially designed browser extension, on the grounds that that it violates that site’s rules against bulk data collection. “Scraping tools, no matter how well-intentioned, are not a permissible means of collecting information from us,” wrote Facebook privacy policy official Allison Hendrix. “We understand the intent behind your tool. However, the browser plugin scrapes information in violation of our terms, which are designed to protect people’s privacy.” If the university doesn’t end the project and delete the data it has collected by Nov. 30, she wrote, it may “be subject to additional enforcement action.”

The Geopolitics of 5G

Efforts by the United States to press allies to shut Chinese vendors out of next generation 5G networks on security grounds appear to be paying off. On October 20, Sweden joined the USA, Australia, the U.K., and France in instituting a ban on using Huawei and ZTE equipment in 5G networks. PTS, the Swedish telecom regulator, said that companies supplying 5G services in the country will have until 2025 to remove any equipment from Huawei and ZTE from their infrastructure networks. PTS added that the decision was based on the advice of Sweden’s military and security services, which apparently described China as “one of the biggest threats against Sweden.” China’s Ministry of Foreign Affairs lashed out, saying that Sweden was using “national security as a pretext to slander Chinese companies, openly oppress Chinese telecom companies, and politicize normal economic cooperation.” The Ministry’s Zhao Lijian warned of “negative consequences.” Ericsson has won contracts from all three major operators in China to supply radio equipment for 5G networks. Retaliation seems imminent. 

Orange and Proximus have picked Nokia to help build 5G networks in Belgium as they drop Huawei amid U.S. pressure to exclude the Chinese firm from supplying key telecoms equipment. The Italian cabinet has vetoed a deal between telecoms group Fastweb and Huawei for 5G core network equipment supplies and is demanding that Fastweb diversify its suppliers. While stopping short of an outright ban on Huawei, the Handelsblatt reports that the German government has raised the bar for participation in 5G networks so that the China-based vendor will unlikely meet the criteria. A draft bill mandating technical tests of components to be used in 5G infrastructure, alongside a ‘political assessment’ of the trustworthiness of manufacturers is being considered and is expected to be cleared in November.

After securing Slovenia’s support for the Clean Network campaign against Chinese technology, the State Department has announced that it had clinched agreements with Slovakia, North Macedonia, Kosovo, and Bulgaria on high-speed wireless network technology. The memorandums make no explicit mention of other countries, but they stress that any new 5G systems should take into account whether the network suppliers are subject “to control by a foreign government,” taking indirect aim at Chinese telecoms giants like Huawei and ZTE.  

In Brazil, the US government is lobbying to ensure that the government chooses Huawei’s rivals. CNBC reports that during a trip to Brasilia on Tuesday, officials from the Export-Import Bank of the United States and U.S. government signed a memorandum of understanding to “identify potential opportunities” for financing up to $1 billion. US officials told reporters at the signing that the money was available for Brazil to buy telecommunications equipment from Huawei’s rivals. 

With the securitization of 5G not subsiding and unable to compete in certain markets or access critical foreign technologies and talent, Huawei is expanding its presence in Russia as well as developing partnerships with Russian universities and tech firms. Huawei is also working on securing supplies for its core telecom infrastructure business despite US sanctions. The Financial Times reports that the company plans to set up a dedicated chip plant in Shanghai as a potential new source for semiconductors after stocks of imported chips Huawei has been accumulating since last year ran out.

Exceptional Access to Encrypted Data

Encryption has always been a politically charged technology, and one that historically has been embroiled in conflicts between individual privacy and national security. A recent “international statement” by the Five Eyes Alliance ((USA, UK, Australia, New Zealand, and Canada) plus “Japan” and “India” calls on technology companies to work with governments to  provide lawful access to the system and build mechanisms that can enable authorities to see content “in a readable and usable format where authorization is lawfully issued.” Startlingly, while other countries have a designated signatory from the government named in the letter, neither India nor Japan mention an authority. 

With this move, law enforcement interests within governments are trying to revive the old argument that their inability to “lawfully access encrypted data and communications poses challenges to law enforcement agencies’ efforts to protect our communities.” Over the past few years we’ve seen several efforts to codify this position in domestic law.  In the UK, the 2016 Investigatory Powers Act put provisions in place to ensure that the government could receive information in an “intelligible form” although it left open exactly how that might occur. Australia’s Telecommunications and Other Legislation Amendment (Assistance and Access) Act 2018 provided authorities the ability to request, require assistance or have a capability be built by a private actor to access encrypted data. Legislators in the United States took up a similar effort with the Lawful Access to Encrypted Data Act proposed in June 2020. 

But it’s not just the Western liberal democracies and friends that are eager to have exceptional access to encrypted data. Researchers at FGV.br flagged several countries (e.g., Brazil, France, Russia, China) where there is existing or proposed law obligating service providers or device manufacturers to provide exceptional access to encrypted data.  With the governments of Japan and India now joining the party we should expect to see legislative proposals there too.

Many civil society advocates have taken up the good fight domestically against this effort to create exceptional access by undoing encryption. The Global Encryption Coalition was launched in May 2020 to promote and defend encryption in key countries and multilateral gatherings where it is under threat. The Internet Governance Project has signed on as a member. 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.