Expressionism
"Regulatory oversight is going to be inevitable," Adam Kinsley, Sky's director of policy, predicted on Tuesday. He was not alone in saying this is the internet's direction of travel, and we shouldn't feel too bad about it. "Regulation is not inherently bad," suggested Facebook's UK public policy manager, Karim Palant.
The occasion was the Westminster eForum's seminar on internet regulation (PDF). The discussion focused on the key question, posed at the outset by digital policy consultant Julian Coles: who is responsible, and for what? Free speech fundamentalists find it easy to condemn anything smacking of censorship. Yet even some of them are demanding proactive removal of some types of content.
Two government initiatives sparked this discussion. The first is the UK's Internet Safety Strategy green paper, published last October. Two aspects grabbed initial attention: a levy on social media companies and age verification for pornography sites, now assigned to the British Board of Film Classification to oversee. But there was always more to pick at, as Evelyn Douek helpfully summarized at Lawfare. Coles' question is fundamental, and 2018 may be its defining moment.
The second, noted by Graham Smith, was raised by the European Commission at the December 2017 Global Internet Forum, and aims to force technology companies to take down extremist content within one to two hours of posting. Smith's description: "...act as detective, informant, arresting officer, prosecutor, defense, judge, jury, and prison warder all at once." Open Rights Group executive director Jim Killock added later that it's unreasonable to expect technology companies to do the right thing perfectly within a set period at scale, making no mistakes.
As Coles said - and as Old Net Curmudgeons remember - the present state of the law was largely set in the mid-to-late 1990s, when the goal of fostering innovation led both the US Congress (via Section 230 of the Communications Decency Act, 1996) and the EU (via the Electronic Commerce Directive, 2000) to hold that ISPs are not liable for the content they carry.
However, those decisions also had precedents of their own. The 1991 US case Cubby v. CompuServe ended in CompuServe's favor, holding it not liable for defamatory content posted to one of its online forums. In 2000, the UK's Godfrey v. Demon Internet successfully applied libel law to Usenet postings, ultimately creating the notice and takedown rules we still live by today. Also crucial in shaping those rules was Scientology's actions in 1994-1995 to remove its top-level secret documents from the internet.
In the simpler landscape when these laws were drafted, the distinction between access providers and content providers was cleaner. Before then, the early online services - CompuServe, AOL, and smaller efforts such as the WELL, CIX, and many others were hybrids - social media platforms by a different name - providing access and a platform for content providers, who curated user postings and chat.
Eventually, when social media were "invented" (Coles's term; more correctly, when everything migrated to the web), today's GAFA (or, in the US, FAANG) inherited that freedom from liability. GAFA/FAANG straddle that briefly sharp boundary between pipes and content like the dead body on the Quebec-Ontario boundary sign in the Canadian film Bon Cop, Bad Cop. The vertical integration that is proceeding apace - Verizon buying AOL and Yahoo!; Comcast buying NBC Universal; BT buying TV sports rights - is setting up the antitrust cases of 2030 and ensuring that the biggest companies - especially Amazon - play many roles in the internet ecosystem. They might be too big for governments to regulate on their own (see also: paying taxes), but public and advertisers' opinions are joining in.
All of this history has shaped the status quo that Kinsley seems to perceive as somewhat unfair when he noted that the same video that is regulated for TV broadcast is not for Facebook streaming. Palant noted that Facebook isn't exactly regulation-free. Contrary to popular belief, he said, many aspects of the industry, such as data and advertising, are already "heavily regulated". The present focus, however, is content, a different matter. It was Smith who explained why change is not simple: "No one is saying the internet is not subject to general law. But if [Kinsley] is suggesting TV-like regulation...where it will end up is applying to newspapers online." The Authority for Television on Demand, active from 2010 to 2015, already tested this, he said, and the Sun newspaper got it struck down. TV broadcasting's regulatory regime was the exception, Smith argued, driven by spectrum scarcity and licensing, neither of which applies to the internet.
New independent Internet Watch Foundation chair Andrew Puddephatt listed five key lessons from the IWF's accumulated 21 years of experience: removing content requires clear legal definitions; independence is essential; human analysts should review takedowns, which have to be automated for reasons of scale; outside independent audits are also necessary; companies should be transparent about their content removal processes.
If there is going to be a regulatory system, this list is a good place to start. So far, it's far from the UK's present system. As Killock explained, PIPCU, CTRIU, and Nominet all make censorship decisions - but transparency, accountability, oversight, and the ability to appeal are lacking.
Illustrations: "Escher background" (from Discarding Images, Boccaccio, "Des cleres et nobles femmes" (French version of "De mulieribus claris"), France ca. 1488-1496, BnF, Français 599, fol. 89v).
Wendy M. Grossman is the 2013 winner of the Enigma Award. Her Web site has an extensive archive of her books, articles, and music, and an archive of earlier columns in this series. Stories about the border wars between cyberspace and real life are posted occasionally during the week at the net.wars Pinboard - or follow on Twitter.