Debates on AI technologies are thus based on the actual or potential demonstration of a risk of bias, discrimination, censorship, or disqualification resulting from computational agents’ entry into society. It is therefore interesting to consider the particularities of the critical discourses articulated in the vast output of expert reports, popularized publications, and works promoting these technolo- gies. The critical dimension of these discourses and spotlighting of controversies surrounding affairs, disruptions and concerns about technological change are most apparent in the media. Between ordi- nary representations and expert reports, the media are the main arena where the promises and risks of AI are put forward. To study them, we carried out a systematic corpus analysis using automatic language processing tools, aimed at identifying the way in which the media frame the problems associated with the deployment of artificial intelligence.