Article 15 of the ECommerce Directive lays down the basic principle that EU Member States cannot impose a general obligation on internet intermediaries to monitor what people say online. We in the UK may have to start worrying for Article 15.
It could easily be overlooked, or even deliberately left behind, when we start the process of converting EU to domestic UK law in preparation for leaving the EU.
Article 15 is a strong candidate for the most significant piece of internet law in the UK and continental Europe. It is the stent that keeps the arteries of the internet open. It prevents the state from turning internet gateways into checkpoints at which the flow of information could be filtered, controlled and blocked.
The principle embodied in Article 15 is currently under pressure: from policymakers within and outside Brussels, from antagonistic business sectors, from the security establishment and potentially from all manner of speech prohibitionists. The common theme is that online intermediaries – ISPs, telecommunications operators, social media platforms – are gatekeepers who can and should be pressed into active service of the protagonists’ various causes.
Article 15 stands in the way of the blunt instrument of compulsory general monitoring and filtering. It does so not for the benefit of commercial platforms and ISPs, but to fulfil the policy aim of protecting the free flow of information and ultimately the freedom of speech of internet users.
Freedom of expression is not just any old policy aim, but a universal value at the heart of human rights – whether we look at Article 19 of the Universal Declaration of Human Rights, Article 10 of the European Convention, Article 11 of the EU Charter, the US First Amendment or the unwritten British historical attachment to freedom of the press. It is particularly precious because, for better or worse, speech reflects our very selves. “Give me the liberty to know, to utter, and to argue freely according to conscience, above all liberties.” (John Milton)
Conversely, freedom of expression has always been threatened by governments whose first instinct is to control. That is one reason why, perhaps more so than for any other human right, defenders of free speech find themselves taking principled stands on the most unattractive ground. “Because if you don’t stand up for the stuff you don’t like, when they come for the stuff you do like, you’ve already lost.” (Neil Gaiman)
The peculiar vice of compelled general monitoring, however, is that we never get to that point. If the filtered and blocked speech doesn’t see the light of day it never gets to be debated, prosecuted, tested, criticised or defended. To some, that may be a virtue not a vice.
Where freedom of speech is concerned, if principle is allowed to take second place to the exigencies of the moment we find ourselves not so much on a slippery slope as in a headlong rush down the Cresta Run. So it is with Article 15. The queue of noble causes on whose behalf we are urged to compel gateways to be gatekeepers – countering copyright infringement, trolling, hate speech, terrorism, pornography, fake news and the rest – stretches round the block.
We defend the right to bad speech for the sake of the good. We understand the impossibility of drawing a bright line between bad and good speech. We regulate bad speech only at the peril of the good. The peril is greater when the regulatory implement of choice is a tool as blunt as general monitoring.
Article 15 lays down a principle that applies across the board, from copyright to terrorism. EU Member States must not impose on internet intermediaries (conduits, hosts and network caches) a general obligation to monitor or actively to seek facts or circumstances indicating illegal activity. Intermediaries cannot be made to snuffle around their systems looking for unlawful activities. Article 15 goes hand in hand with the Directive’s liability shields under which conduits, hosts and network caches have various degrees of protection from criminal and civil liability for the activities of their users.
It is only too easy for policymakers to point the finger at intermediaries and demand that they do more to control the unpleasant and sometimes illegal things that people do on their systems. Policymakers see intermediaries as points of least cost enforcement: it is more efficient to enforce at a chokepoint than to chase tens of thousands of individual wrongdoers. The theory is made explicit in Recital (59) of the EU Copyright in the Information Society Directive:
“In the digital environment, in particular, the services of intermediaries may increasingly be used by third parties for infringing activities. In many cases such intermediaries are best placed to bring such infringing activities to an end.”
Mr Justice Arnold in Cartier explained the policy that underlies Recital (59):
“As can be seen from recital (59) to the Information Society Directive, the economic logic of granting injunctions against intermediaries such as ISPs is that they are the “lowest cost avoiders” of infringement. That is to say, it is economically more efficient to require intermediaries to take action to prevent infringement occurring via their services than it is to require rightholders to take action directly against infringers. Whether that is correct as a matter of economics is not for me to judge. Nor is it for me to judge whether it is good policy in other ways. That judgement has already been made by the legislators …”
At the same time, Article 15 of the ECommerce Directive constrains the breadth of injunctions that courts can grant against intermediaries under the Copyright and Enforcement Directives. The effect of Article 15 can be seen in the ECJ decisions of SABAM v Scarlet and SABAM v Netlog prohibiting content filtering injunctions, and in Arnold J’s Cartier judgment itself:
“If ISPs could be required to block websites without having actual knowledge of infringing activity, that would be tantamount to a general obligation to monitor.”
But if intermediaries are best placed to stop infringement, why should Article 15 constrain what can be imposed on them? Why shouldn’t the intermediaries be required to monitor?
The only sense in which intermediaries could be seen as best placed is that, since users’ communications flow through their systems, they have the potential to be technical chokepoints. In every other respect intermediaries are poorly placed to make decisions on legality of content and thus on what to block.
Intermediary enforcement risks exaggerating the ease with which unlawful behaviour can be identified, often assuming that illegal content is identifiable simply by looking at it. In relatively few categories is illegality manifest. Legality is predominantly a matter of factual investigation and judgement. That is why it is preferable to have independent courts ruling on matters of illegality rather than compelling private platforms to attempt it and have them overblock out of fear of liability or sanctions.
A too narrowly focused cost analysis tends to underplay or even ignore the negative externalities and unintended consequences of compelling gateways to act as gatekeepers. It excludes any broader implications of reinforcing chokepoints, the creation of a climate in which playing gatekeeper on behalf of the state and its proxies becomes the norm. In a broader context the least cost enforcer may turn out to be highest cost.
Notice-based intermediary liability systems result in material being removed before a court determines whether it is unlawful. That already carries a risk of overcautious blocking or removal. Compelled proactive monitoring and filtering, since it blocks information about which no complaint has been made, moves the scale of risk to another level. It is akin to prior restraint on a grand scale, conducted not by courts after hearing evidence but by private entities made to act as investigator, prosecutor, judge, jury and executioner.
Our aversion to prior restraint reflects also that the public are sometimes well served by the airing of something that at first blush might appear to be against the strict letter of the law. Speech may be rendered lawful by a public interest defence, or by fundamental freedom of speech considerations. Or a court might decide that even though unlawful the appropriate remedy is damages but not removal. Legality of speech, even in areas such as copyright, can be a heavily nuanced matter. Proactive general monitoring obligations allow for no such subtlety.
Some may argue that in modern times the quid pro quo for living with freedom of speech has been that speech is generally mediated through professional, responsible editors. And that we need to put that genie back in the bottle by converting online intermediaries into editors and publishers, responsible for what other people say on their platforms.
Never mind whether that could be achieved, the argument misunderstands the nature of freedom of expression. The great advance of the internet has been to bring about something akin to the golden age of pamphleteering, freeing mass individual speech from the grip of the mass media. District Judge Dalzell was right when, in ACLU v Reno, he said:
“As the most participatory form of mass speech yet developed, the internet deserves the highest protection from governmental intrusion.”
The US Supreme Court in the same case said:
“Through the use of chat rooms, any person with a phone line can become a town crier with a voice that resonates farther than it could from any soapbox. Through the use of Web pages, mail exploders, and newsgroups, the same individual can become a pamphleteer.”
Those quotations were from 1996 and 1997. They are, if anything, more relevant now. Individual, unmediated speech deserves more, not less, protection than the traditional press.
It may be discomfiting that the kind of vitriol that used to be confined to Speaker’s Corner can now reach an audience of millions. But freedom of individual speech was never something only to be tolerated as a tourist curiosity, or indulged as long as it was hidden away in a pub saloon bar. Nor, as we know from the ECtHR decision in Handyside, is freedom of expression confined to that which would not offend in a genteel drawing room.
Article 19 of the 1948 Universal Convention on Human Rights is not predicated on the assumption of mediated speech. It articulates an individual, personal right that transcends place, time and medium and could have been written with the internet in mind:
“Everyone has the right to freedom of opinion and expression; this right includes freedom to hold opinions without interference and to seek, receive and impart information and ideas through any media and regardless of frontiers.”
Article 15 stands squarely in the path of compelling mediated speech through the instrument of general monitoring. So why might it be vulnerable to being overlooked in the Brexit process?
Caveat: what follows is based on the existing Conservative government’s plans for a Great Repeal Bill and is subject to the outcome of the General Election.
Article 15 and Brexit
Article 15 operates at several levels. If a Member State were to legislate in breach of Article 15, a national court would be obliged to disapply the legislation. So it acts as a powerful constraint on Member States at the policy and legislative level. As we have seen it also constrains Member States’ courts. They cannot issue an order that would impose a general monitoring obligation. They cannot interpret a domestic statute or develop a common law obligation in a way that is contrary to Article 15. Regulatory and enforcement bodies are similarly constrained.
The Brexit starting point is that the incumbent government has committed to continuing existing EU law through the Great Repeal Bill. Theresa May says in the introduction to the Great Repeal Bill White Paper:
“Our decision to convert the ‘acquis’ – the body of European legislation – into UK law at the moment we repeal the European Communities Act is an essential part of this plan.
This approach will provide maximum certainty as we leave the EU. The same rules and laws will apply on the day after exit as on the day before. It will then be for democratically elected representatives in the UK to decide on any changes to that law, after full scrutiny and proper debate.
… Th[e Great Repeal] Bill will, wherever practical and appropriate, convert EU law into UK law from the day we leave so that we can make the right decisions in the national interest at a time that we choose.”
On that basis Article 15 ought to be continued post-Brexit. However there is a technical problem. Although it is in a Directive, and so was required to be implemented in UK law, the text of Article 15 appears nowhere in UK domestic legislation. Depending on how the proposed Great Repeal Bill is drafted, Article 15 may have to be specifically written in to UK legislation in order to continue post-Brexit.
The White Paper recognises the need to write EU Regulations into domestic law, but appears to assume that since a Directive will already have been implemented in UK domestic law it just needs to be preserved post-Brexit:
“• the Bill will convert directly-applicable EU law (EU regulations) into UK law;
• it will preserve all the laws we have made in the UK to implement our EU obligations”
Article 15 could run the risk of falling between the cracks.
In any event the desirability of continuing Article 15 may not be universally accepted. UK Music, in its ‘Music 2017 Manifesto’, has noted the opportunity that Brexit presents to ‘place responsibility on internet service providers and require them to have a duty of care for copyright protected music’. If that implies proactive monitoring it would put Article 15 in question. Where one industry leads others may follow. A government interested for its own purposes in turning the screw on intermediaries might not welcome the impediment of Article 15. It might be tempted to invoke the ‘wherever practicable and appropriate’ White Paper qualification on continuation of existing EU law.
“Freedom of expression is not self-perpetuating, but rather has to be maintained through the constant vigilance of those who care about it.” So said Index on Censorship in 1972. The run-up to Brexit may be a time for especial vigilance.
This post originally appeared on the Cyberleagle blog and is reproduced with permission and thanks.