by Jan Christopher Kalbhenn, LL.M. (Amsterdam)
Background: The rise of Social Media
Obviously one of the most pressing challenges for media regulators is the increasing use of algorithmic-driven news aggregators and the growing dominance and gatekeeper position of big platforms, especially social media platforms. The Internet serves as a space for public communication and opinion formation. Public discourse, advertisement and election campaigns are increasingly taking place online. For example, the reach of the climate movement “Fridays for Future” in social media shows that the Internet has become significantly more relevant for opinion forming. According to the Digital News Report 2019 in Germany 48% of the people aged between 18-24 say that their most important gateway to news is via one of the algorithmic-driven sources (Social Media, News Aggregators, Search-Engine). In the overall population the score is at 32%.
The threat to diversity of opinion
This structural change not only offers opportunities, but also poses certain risks for the diversity of opinion which is essential for a democracy. For diversity of opinion, it is crucial that as many diverse perspectives as possible are included in the free individual and public opinion-forming process. Intermediaries generally increase the supply of information. Nevertheless, a multitude of information offerings on the Internet does not necessarily guarantee diversity of opinion. Today, the question of the relevance of news and information is decided by its visibility and findability on the internet. The selection and control of content by intermediaries and aggregators is of central importance in this context. It is a characteristic of search engines and social networks that they filter, personalise and present content to users based on personal data they have collected on them. To serve the overarching business model algorithms are programmed to ensure that the users spend as much time as possible on their respective platforms. However, this can be problematic. The criteria codified in the algorithms alone decide which information offers can be perceived and which cannot. It often remains unclear on which basis the content is selected by the algorithms. This lack of transparency makes public opinion formation potentially susceptible to manipulation. For example did TikTok lately curb the reach for people with disabilities? At the very least, the public discourse is privatised according to the logic of an attention-based business model. The public discourse then is subject to the rules of private entities acting on behalf of their shareholders.
The need of a new regulation
To summarise the challenges for democracy posed by Social Media one can paraphrase the German Constitutional Court. In July 2018 it had to rule on questions regarding the financing of the public service media. Without being asked the court gave its pronounced opinion on the digitalisation of the media: “The digitalisation of the media and in particular the focus on Internet networks and platforms, including social media, fosters tendencies of concentration and monopolisation in respect of content provider, disseminators and intermediaries.” It continues: “These services do not aim to reflect diverse opinions; rather, they are tailored to the rationale of a business model that aims to maximise the time users spend on a website” (English version here).
The new Interstate Media Treaty
In Germany, so far media regulation in the sense of promoting diversity more or less only knows linear TV. The German Media concentration law is practically inapplicable in the digital world. The positive diversity rules for broadcasters are not applicable on intermediaries nor are the must-carry rules for (cable) platforms. The network enforcement law known across German borders is the first law to regulate social networks. However, it does not directly aim at diversity of opinion but at securing a free discourse. It obliges social networks to speedily delete illegal content (e.g hate speech). So that media law meets the digital challenge, in Germany, there will soon be a major change in the media regulation. Media law will for the first time include so called media intermediaries in the regulatory framework. The new Interstate Media Treaty contains countless other amendments. It implements the AVMS-directive and includes video sharing platforms in the regulatory framework. Also new is an extension of the journalistic due diligence obligations in the online sector, including a new co-regulatory mechanism. The Interstate Media Treaty – Medienstaatsvertrag – will most likely enter into force in September 2020. It must be ratified by all state parliaments, as media law in Germany is a matter of the Länder.
Who is addressed by the new diversity rules?
Within the framework of the new regulation, the circle of addressees is expanded to so-called ‘Media Intermediaries’. The definition on Media Intermediary in the new treaty in section 2 no 16 reads as follows: “Any tele media which aggregates, selects and presents also journalistic and editorial offers of third parties in a generally accessible way without combining them into a complete offer”. The new rules do not apply to Media Intermediaries who a) reach less than one million users in the federal territory per month in any case; b) are specialised in the aggregation, selection and presentation of content related to products or services; c) serve exclusively private or family purposes. The main subjects of application of the rules are therefore social networks, search engines and news aggregators: Facebook, Twitter, Google and many more. They will all be subject to media law regulation in Germany for the first time.
What are the new ‘diversity obligations’?
There are two main obligations for the Media Intermediaries focusing on diversity. First the transparency obligations and second the non-discrimination obligation.
- Transparency obligations
Section 93 states that in ‘order to ensure diversity of opinion’, providers of Media Intermediaries shall keep some information easily recognisable, directly accessible and permanently available. Firstly, the criteria which determine the accessibility of a content to a media intermediary and its remaining, and secondly, the central criteria of aggregation, selection and presentation of content and their emphasis, including information on the functioning of the algorithms used. All information have to be in understandable language. Any changes in these criteria shall be immediately notified in the same way. Therefore, in the end this may be what is called “algorithmic transparency”. Here it will have to be shown how this can be implemented in practice. The state media authorities will play a central role in this. Not only in the monitoring process. First of all, they are commissioned to concretise the legal requirements in joint guidelines. In view of the wording, which is open to interpretation, this might be helpful. However, there must be no deviation from the principle that the essential decisions must be made by the legislator.
There is another obligation regarding intermediary transparency. It is about the labelling of Social Bots. According to the new treaty social networks shall ensure that tele media on their platform indicate the circumstance of automation. The shared content or message must be accompanied or preceded by a clearly legible notice declaring that it was created and sent automatically using a computer program controlling the user account. These provisions on social bots do not impose certain diversity obligations on intermediaries, neither positively nor negatively, with regard to media content, but are intended to ensure a minimum standard of equal opportunities for communication.
- Non-discrimination obligations
In section 94 the new law states that in order to ensure diversity of opinion, media intermediaries may not, either directly or indirectly, unreasonably obstruct journalistic and editorial offers on whose perceptibility they potentially have a major influence or treat them differently without objectively justified reason. According to the new law, discrimination takes place in particular if there is a deliberate and targeted deviation from the general rules of aggregation, selection and presentation within the meaning in favour of or at the expense of a specific content. In the future, one of the decisive factors will be how the undefined legal term “on whose perceptibility they potentially have a major influence” is interpreted. To this end, demands are already being made in the debate to remove this restriction. Again, the state media authorities will play a central role by its concretization of the new requirements in joint guidelines.
Who will be in charge?
Only the provider of the journalistic and editorial content concerned may claim a violation of the non-discrimination obligations. Enforcement lies within the power of the competent 14 state media authority. These are organised on a federal and state-funded basis and they control compliance with the Interstate Media Treaty. The State Media Authorities are to develop common guidelines to specify the new rules for intermediaries. On the one hand, this makes sense in order to make the criteria of the law manageable, some of which require interpretation. On the other hand, the constitutional principle of essentiality must be observed here. This states that the legislator (and not, as in this case, the state media authorities) must make the essential decisions.
The state media authorities have far-reaching rights of information and investigative powers vis-à-vis the media intermediaries. Media intermediaries may be sanctioned if they violate their obligations. Firstly, some infringements are defined as misdemeanours and can be punished with fines of up to EUR 500,000. In addition, the state media authorities are also empowered to take the ‘necessary measures’ according to the general clause of Section 111.
How is the new law to be assessed?
The Interstate State Media Treaty is the first attempt to adapt the guarantee of diversity of opinion to the digital space. By its very nature it is difficult to predict the success of a regulation as complex as the Interstate Media Treaty. However, it is relatively easy to predict that the current regulation (Rundfunkstaatsvertrag/Interstate Broadcasting Treaty) would reach its limits sooner rather than later. The main reason for this is that a large number of purely digital phenomena are currently being either completely ignored or at least only helpfully regulated. Nevertheless, the change in media use and the convergence of the media makes it necessary to look at new distribution channels. Consequently, the new treaty also includes virtual gatekeepers in the regulatory framework.
It remains to be seen whether this new regulatory model will have resounding success. In any case, it is a courageous and important first step to tackle the challenges of the digital media era. Initial findings will emerge and on the basis of these findings, further steps can be taken. This is particularly important, as the State Media Authorities will be confronted for the first time with digitally dominant global corporations and will have to control and, if necessary, sanction them. International observers will take a close look to see if the new rules work. Or where there is room for improvement. The German debate is already calling for further measures. For example, the report of the Data Ethics Commission of the Federal Ministry of the Interior calls for stricter rules to apply to media intermediaries who act as gatekeepers for democracy. According to the Commission to reduce risks to democracy, ex-ante controls (e.g. in the form of a licensing procedure) should also be considered. A further proposal of the Commission is to oblige gatekeepers to ‘use recommendation algorithms that allow users (at least as an additional option) access to unbiased and balanced information that embodies pluralism of opinion’. However, the debate has just begun.
In any case, procedures for access to the data the platforms collect by distributing media companies’ content would be required to balance the level playing field between the media and the platforms (see https://search.coe.int/cm/pages/result_details.aspx?objectid=090000168092dd4d). It is to be hoped that contributions will also be made at a European level, within the framework of the competencies, to secure media diversity and the diversity of opinions in the digital era. The big shot would certainly be a digital European media platform oriented towards the public interest – Public Open Space (e.g. the proposal of Hillje, Plattform Europa, 2019). This platform would work with algorithmic (recommendation-) systems based on democratic values and criteria. Recommendations should then have a constructive bias. Cognitive scientists would have to be involved. For this purpose, a more cooperative media order would first be necessary.
By Jan Christopher Kalbhenn, LL.M. (Amsterdam)
Managing Director of the Institute for Information, Telecommunications and Media Law at the University of Münster and doctoral candidate at the chair of Prof. Bernd Holznagel.
The article is based on the manuscript of a speech the author gave at the EUI CMPF in Florence on 31 January 2020.