In this second post on the Swedish proposal for implementation of Article 17, I look at provisions explicitly concerning users of services. Whilst certain user safeguards are already embedded in the liability regime (Part 1), the Swedish Ministry of Justice goes the extra mile and proposes to comprehensively address user interests in the text of the Copyright Act.
3. Dessert: User rights proper with a few cherries on top
3.1 A mandatory right to make non-infringing content available to the public
If you like legal theory you probably think of Axel Hägerström, Alf Ross, Vilhelm Lundstedt or Karl Olivercrona every time you hear Scandinavia. Or if you really know your Scandinavian legal theory: the Uppsala school of legal realism. The legacy of this school of thought, that dominated 20th century Sweden, is an entrenched scepticism towards the notion of rights, especially at constitutional level, such as fundamental rights. In what could be seen as a desire to enter a new path and an attempt to treat seriously the CJEU’s 2019 judgments in Spiegel Online and Funke Medien (in which the Court assessed with the help of the EU Charter that copyright exceptions confer rights on users), the Swedish Ministry of Justice proposes the introduction in section 52p of a user right to make available content for the purpose of quotation, criticism, review, caricature, parody, and pastiche (cf. Article 17(7)), and any other non-infringing use. Accordingly, the proposed right is intended to incorporate the entire catalogue of copyright exceptions, the Swedish freie Benutzung provision in section 4, uses of unprotected content (for example which are in the public domain) and licensed uses (Ds 2021:30, pp. 298-299).
Thus, not only is a service obliged to ensure that lawful content is not disabled, as follows from the proposed obligation in section 52o (see Part 1), but users actually have a positive right to make non-infringing content available. More than being an empty statement, the newly proposed right entrenches the user safeguards embedded in the liability regime and is actually intended to prevent a service from disabling access to lawful content with reference to the safe harbour; i.e. by citing provisions which are inherently intended to protect copyright (Ds 2021:30, p. 299; however such content may be disabled on other legal grounds than copyright if the circumstances call for it).
On top of it all, terms contracting out the new user right(s) are proposed to be null and void (sv. ogiltig).
But there is more.
Under the proposed section 52r, a service is liable to compensate harm that a user suffers as a result of intentional or negligent disregard of the obligations imposed in section 52o (lack of effective routines, automatic blocking of lawful content or that is in the grey area, and failure to notify users when access to content has been disabled).
Moreover, on the basis of section 52s a user (or an organisation representing users) is able to initiate proceedings before a court concerning the issuing of an injunction on penalty of a fine against a service to take corrective action when the service has disregarded its obligations in sections 52o, 52q first paragraph (lack of complaints procedure), and 52q third paragraph (non-expeditious complaint handling, non-restoration of content when the rightholder has not duly justified the request to disable access, and lack of human review in case of a decision to not restore access to content).
In sum, the vision of the Ministry of Justice is that Article 17 takes the form of an updated hosting safe harbour that this time explicitly accounts for users of services.
4. That wafer-thin after-dinner mint à la Mr Creosote?
With all the delays in the legislative process, the question then is whether Sweden has failed to implement Article 17 on time. Of course the DSM Directive also contains other important provisions and those have not been implemented yet. However, when it comes to platform liability, ever since 1998 (you read that right; the 1990s) these and other kinds of services have been under a statutory duty to remove or otherwise prevent dissemination of infringing content via their services pursuant to section 5 BBS Act (“Bulletin Board System”; if you know your internet). Combined with the voluminous case law on the communication to the public right in Article 3 InfoSoc Directive, including the clarification in YouTube/Cyanado, the recognition of copyright exceptions as conferring rights on users, as well as the existence of a general extended collective licensing scheme under section 42h Copyright Act since 2013, it appears to me that we already have a version of Article 17 in place which, one could argue, is perfectly fine to achieve the result pursued by this provision of the DSM Directive (cf. Article 288 TFEU). All things taken together, the application of this old dinosaur does not result in a very different outcome than what the Ministry proposes today, especially (and interestingly!) considering that the BBS Act similarly does not extend to linking (Prop. 1997/98:15, p. 17). The caveat to this conclusion is that the BBS Act does not contain any redress mechanism for users.
I’m not fan of the so-called homing tendency in a European copyright climate, but nowhere does the Ministry’s proposal attempt to make use of what is already in place and arguably in a far less convoluted manner. This last aspect is my main critique of the Ministry’s proposal that I otherwise find difficult to criticise. The BBS Act, with all its 1990s simplicity and charismatic terminology leaning on communication theory (it refers to senders and receivers of messages, which are defined broadly) requires of the operator of a service to prevent obvious infringements. It does not require a review of every incoming message but does require exercise of such control over the service which is reasonably required having regard to the nature of the service (Prop. 1997/98:15, p. 15). This is the essential core of the Article 17 regime following the Commission’s guidance and what the Ministry also proposes now (with a few cherries on top). Moreover, compatibility of the BBS Act with Article 15 E-Commerce Directive was assessed when the latter was being implemented in Sweden with the result that the BBS Act is compatible because it does not require active investigation of facts or circumstances which can imply unlawful activity (Prop. 2001/02:150, p. 100).
So why not keep it as it is and merely serve the dessert (a Swedish fika, if you like) by implementing user safeguards to help preserve the exercise of fundamental rights in the digital environment and in that way circumvent unnecessary complexity and overregulation? Having three frameworks (Article 17 services, non-Article 17 services, and the BBS Act on the side) will become difficult to handle when copyright law is developing fast and in all directions. With ever stronger need for flexibility in the system, we might actually be better off if we use the Ministry’s reasoned proposal for Article 17 as a guide for how to correctly construe and apply the BBS Act a quarter of a century later in copyright matters, possibly with some tweaks.
The most fascinating thing is that during this entire time there have hardly been any copyright cases centring on the BBS Act; despite the fact that a copyright case was the very reason for its adoption (NJA 1996 s. 79, the BBS case). So it’s either very effective or, since liability under the Act is subsidiary to liability under copyright law, maybe there never really has been any problem with preventing infringements online in Sweden. Either way, I would hesitate to place the BBS Act in the attic – with its beautiful simplicity it also covers other contemporary information law issues, for instance hate speech, terrorist content, or interferences with privacy, making therefore Sweden also (already) have a mini-Digital Services Act in place (I know, it’s going to be a Regulation..).