An Alien in the Newsroom: AI Anxiety in European and American Newspapers

Abstract

The media portrayal of artificial intelligence (AI) directly impacts how audiences conceptualize this technology and, therefore, its use, development, and regulation. This study aims to measure a key aspect of this problem: the feeling of AI anxiety conveyed by news outlets that represent this technology as a sort of “alien” that is autonomous, opaque, and independent of humans. To do so, we build an AI anxiety index based on principal component analysis (PCA) and apply it to a corpus of headlines (n = 1682) about AI published before and after the launch of ChatGPT in ten newspapers: The New York Times, The Guardian, El País, Le Monde, Frankfurter Allgemeine Zeitung, San Francisco Chronicle, Manchester Evening News, La Voz de Galicia, Ouest France, and Münchner Merkur. The results show that ChatGPT not only boosted the number of AI headlines × 5.16) but also reduced positive sentiments (−26.46%) and increased negatives (58.84%). The AI anxiety index also grew (10.59%), albeit driven by regional media (61.41%), while it fell in national media (−6.82%). Finally, the discussion of the variables that compose the index reveals the opportunities and challenges faced by national and regional media in avoiding the feeling of AI anxiety.