GENERATORI DI LINGUAGGIO, DISINFORMAZIONE E PSEUDO-VERITÀ

Language generators are AI systems capable of creating and processing
knowledge that bears a resemblance to natural language. Their capabilities make
them the ideal tools for use in disinformation practices. At the same time, they are
a technology that is set to change the way society accesses, processes and uses information. This prompts the emergence of intricate legal concerns, particularly
given the potential for AI systems to produce erroneous, false, or seemingly accurate outputs. The present essay addresses these issues in the context of the regulations recently introduced by the European legislator, taking into account the roles
and responsibilities of the various socio-economic actors involved, not only the
providers of language generators, but also those who use them and, finally, the end
users of their output. This study focuses on the utilisation of this technology in the
context of digital services, encompassing both those provided by intermediary service providers as outlined in the Digital Services Act and language generation services directed towards a broad and undifferentiated audience. The latter, in particular, give rise to issues concerning consumer rights when users are natural persons who use the output for purposes unrelated to any professional activities they may carry out.