Last week, European Digital Rights attended the second annual Stockholm Internet Forum which focused on two main themes: Internet Freedom and Security and Internet Freedom and Development. A novelty this year were the Unconference sessions.

The Unconference organised by EDRi set itself the task of establishing an initial set of basic criteria for assessing the appropriateness of intermediaries taking voluntary measures to achieve public policy or other goals.

This summary does not bind any of the participants in the very energetic and collaborative discussions and is meant to serve as a contribution to the growing debate surrounding this controversial topic.

The task at hand was summarised by one participant as the “achievement of public policy or other goals through voluntary or mandatory measures by internet companies, including through their ‘accidental constitutions’ (their terms of service) which de facto regulate their ‘quasi public spaces’”.

An industry representative stressed at the outset that such measures can be valuable, offering a flexible and rapid opportunity to address issues that arise, although it was recognised that this is suitable in some circumstances but not in others. The discussion therefore sought to establish the characteristics of situations where such approaches are likely to be effective, proportionate and in line with wider public policy objectives like respect for freedom of communication, the rule of law and democracy.

Criterion 1: Is the process internal or external to the intermediary?

Broadly speaking, the more the process is internal to the intermediary, the more likely it is that the measure will be effective. Internal motivations and implementation (solving something that is a direct problem for the intermediary through internal processes) are likely to lead to measures that achieve the public policy objective more efficiently than external motivations (such as avoidance of legislation) and external implementation (such as penalising customers or third parties for alleged infringements of law).

Criterion 2: Are there vested interests on the part of the intermediary?

Voluntary or mandatory interventions by intermediaries to achieve public policy objectives may be supported or initiated by companies as a way of achieving a competitive advantage. This can result in them taking punitive actions against competitors or by lobbying for measures which only incumbents (due to economies of scale) can easily implement. This can lead to unintended (economic and/or societal) consequences that are disproportionate to the public policy objective being pursued.

Criterion 3: How competitive is the market?

There are cases, online marketplaces for example, where the nature of the service and competitive environment may be adequate to ensure that voluntary interventions by the service provider would not result in any significant competitive or practical impact on the user whose activities are restricted. The real choices available to the subject of the measures in question are therefore of importance in assessing the value of a voluntary intervention.

Criterion 4: What is the (public) policy objective being pursued?

Is the intervention seeking to address the business or public relations concerns of the company (through a ban on content which is not illegal, for example) or to enforce a specific law? Interventions to implement a democratically agreed law clearly have more legitimacy than other measures.

Criterion 5: Whose law is being implemented?

Is the intermediary implementing (voluntarily or otherwise) a law that has been democratically approved in the country in question or (also) in countries outside the jurisdiction that adopted the law? This question is crucial for the democratic legitimacy of the activity.

Criterion 6: Are there regional variations of the impact of the measures?

Currently, certain social networks are given preferential treatment by mobile operators in some countries. As a result, the choices available to Internet users in those areas are significantly narrower, which changes the assessment of the proportionality of any measure. Voluntary interventions, especially to regulate activities which are not illegal, would be particularly inappropriate in such circumstances.

Criterion 7: What is the responsibility of the intermediary for its interventions and does the citizen have a right of redress?

Is there a practical legal way of making the company responsible for the impact of the measures it takes (particularly if voluntary) and is effective redress available for the user? If not, this creates a situation where the intermediary has significant power but limited or no responsibility. It is clear that such situations should be avoided.

Criterion 8: What is the collateral damage for liability exceptions?

The “safe harbour” protections offered to Internet intermediaries have been crucial in the development of an open Internet, protecting free speech against arbitrary, defensive measures being taken by intermediaries. There is therefore a significant danger of voluntary measures being (mis)used to reduce liability protections.

For more information on this topic see our booklet: “The Slide from Self-Regulation to corporate censorship”: