SOCIETÀ ITALIANA DI DIRITTO ED ECONOMIA
Massimiliano Vatiero (Università degli Studi di Trento, and Università della Svizzera italiana)
Abstract
Non-traditional information providers, such as digital platforms (e.g., Facebook, X) and AI-based products (e.g., ChatGPT), have improved access to information and increased the number and diversity of information sources. However, there are social costs as well. First, people are exposed to a daily dose of false or misleading content. Digital platforms foster polarization and spread fake news and stereotypes (e.g., Allcott and Gentzkow 2017, Sunstein 2017, Vosoughi et al. 2018). For instance, news stories on Twitter and posts on Facebook during and following the 2016 US presidential elections played a key role in driving the viral spread of misinformation (Shao et al. 2018), which even led to violent accidents (e.g., the Capitol Hill siege). Similar concerns have been raised about AI’s ability to generate deepfakes (cf. Nestor et al. 2024). For instance, Donald Trump supporters have been utilizing AI to produce and distribute fabricated images of Black voters endorsing the Republican party, contributing to a misinformation trend (cf. Spring 2024). Moreover, Omiye et al. (2023) found racial bias when AI-based products are prompted to respond to medical questions. Finally, misinformation from digital platforms and AI-based products can (and often does) even multiply each other. For instance, in January 2024, sexually explicit, AI-generated (false!) images purportedly depicting Taylor Swift surfaced on X. In a political context, shortly before the election, an AI-manipulated and false audio clip emerged on Facebook, capturing Michal Šimečka, the leader of the Progressive Slovakia party, and a journalist, discussing illicit election strategies. Since the clip was released during a pre-election quiet period, when media and politicians’ commentary were restricted, the clip’s dissemination was not easily contested. Ultimately, the affected party, Progressive Slovakia, lost by a slim margin to SMER, one of the opposition parties (see Nestor et al. 2024).
Departing from the so-called “marketplace of ideas” theory, this paper examines (i) whether and how antitrust and competition policies may ensure the supply of true information, (ii) whether the provision of the truth, as the provision of a merit good, requires a derogation of market competition instead.