The emergence of generative artificial intelligence brings back to the table the question of remuneration for content producers of all kinds. If the amounts involved are nothing comparable (for the moment), they clearly show the importance of this question.
If generative AI promises considerable benefits, it can also be frightening, like any potentially disruptive technology. In the cultural and creative sectors, it raises fears of the replacement of humans and the loss of authenticity of creations or infringement of copyright. The first conflicts are emerging in the world between owners of intellectual works and artificial intelligence players, around respect for intellectual property. This legal and ethical issue is coupled with the economic issue of the distribution of income generated by AI.
It is not certain that the specific framework of copyright, which is based on the representation or reproduction of a particular work, is at stake in these conflicts. On the other hand, it is certain that there is no simple way to assess the improbable contribution of a particular work to the creation of a work generated by artificial intelligence, while copyright is based on the principle of proportional remuneration (to the income generated by the work).
[Déjà plus de 120 000 abonnements aux newsletters The Conversation. Et vous ? Abonnez-vous aujourd’hui pour mieux comprendre les grands enjeux du monde.]
An asymmetrical relationship
These new oppositions are in line with the numerous conflicts that have arisen for years in the relationships between platforms and holders of protected content. Platforms have taken a considerable place in the economy of cultural industries: Spotify, YouTube or TikTok in music, Netflix or Apple in audiovisual, Steam in video games, Google or Facebook in the press…
They transform sectors, bring new modes of consumption, weigh on the visibility of works, influence their creation, structure new power relationships. On the one hand, they are vital for many creations of the mind in these areas; on the other, they need these contents, but in an asymmetrical relationship. A song banned from Spotify can see its prosperity called into question when the opposite is not true: at most, the absence of a major artist can result in a marginal churn rate to the benefit of its competitors.
A Nobel Prize for the platform economy
The platform economy has been the subject of numerous studies. Their properties of two-sided markets, the demonstration of which was awarded a Nobel Prize to Jean Tirole in 2014, put them at the center of two markets – two sides – which reinforce each other: the more a platform highlights content, the larger its audience; the larger its audience, the more it attracts content owners. This leads to a dynamic of concentration and one-sided subsidizing strategies. However, the works which have theorized the economy of platforms have only studied the relationship to “content” in a piecemeal manner. They do not differentiate between content types. However, some of them, those that are sometimes referred to as “premium”, such as live sports competitions, have a singular status which distinguishes them from the mass of others. And they do not provide an answer to the question of the value brought by such content to the economy of a platform.
Read more: The global AI scene: between regulations and soft power
However, it is this question which is at stake behind the conflicts between platforms and content providers, and those which are beginning to appear between AI operators and content holders. Behind it is the need for a new framework. The numerous conflicts that have unfolded over the years clearly show that yesterday’s tools are no longer suitable for tackling these issues.
Difficult to know what brings what
The example of the press is edifying. For years, conflicts have opposed press publishers in many countries, who demand remuneration for the resumption of their content, to platforms, mainly Google and Meta. Google references the press among many other contents. Highlighting press content improves the quality of the results offered by Google for a query, and contributes to the value of the platform’s services which spontaneously highlight them. But this value is not identifiable because Google’s economic model is systemic: it is all the content referenced which gives value to its services, some highlights are paid for by advertising and others are not. (organic or sponsored results), certain services are not directly linked to an advertising market (Google News for example) and all of Google’s services generate data which makes both the search service and addressing more effective. advertisements.
Under these conditions, it is difficult to isolate the value of a particular content. Furthermore, content may be presented in the form of a hypertext link (which therefore refers to the site of the protected content holder) or a summary, which does not refer to the site of the initial producer of the information. Depending on the Internet user’s action, he will therefore see elements of a press article while remaining in the Google environment or will be sent to the press publisher’s site. In one case, Google can be considered as an operator of content (but on a page among many others, without it being possible to know whether it has been read) or as providing readers to its publisher. As a prescriber, he brings value to the content; as an operator, he earns from it.
A new paradigm
At the macro level, the press economy suffered from the arrival of these actors who took advantage of this content. But at the level of content or a publisher, what is its contribution to the value produced by Google or Meta? Depending on the case, these conflicts take different forms and do not converge towards solutions that would become predictable.
Depending on the country, agreements are reached, then called into question, different jurisdictions are seized, the balance of power takes over (with situations in which the platform stops referencing press content). And the courts seized do not comment on the level of value sharing. This uncertainty and instability is due to a paradigm shift. In the platform economy, content, or even a category of content, no longer has an identifiable contribution. Given the importance of platforms in the economy of cultural industries, inventing a framework adapted to the novelty of these situations is urgent.
