Botlitica: A generative AI-based tool to assist journalists in navigating political propaganda campaigns

Musi, E ORCID: 0000-0003-2431-455X, Garcia Aguila, EE and Federico, L (2024) Botlitica: A generative AI-based tool to assist journalists in navigating political propaganda campaigns Studies in Communication Sciences, 24 (1). pp. 161-169. ISSN 1424-4896, 2296-4150

PDF 11._Musi_et_al._161-169 (1).pdf - Other Download (441kB) | Preview

Abstract

The hype on generative AI has raised concerns about the spread of disinformation but also opened up new opportunities for hybrid journalism. The proliferation of political propaganda campaigns spread across digital media during election periods constitutes a challenge for journalists who struggle to exercise information gatekeeping. AI-based tools can in principle help, but journalists are resistant to using them since they are sceptical about their compliance with news values and their scarce user-friendliness. To cope with such issues, we present Botlitica, a GPT3-based chatbot able to answer users’ questions according to the information shared on different social media by a given political party which is, thus, embodied by the AI agent in the conversation. The back-end and front-design of the chatbot are devised to privilege transparency. We report the results of a preliminary evaluation of Botlitica which show that the tool fastens journalists’ capabilities to navigate propaganda campaigns inducing them to exercise critical thinking.

Item Type:	Article
Uncontrolled Keywords:	4701 Communication and Media Studies, 47 Language, Communication and Culture, Machine Learning and Artificial Intelligence
Divisions:	Faculty of Humanities & Social Sciences > School of the Arts
Depositing User:	Symplectic Admin
Date Deposited:	10 May 2024 10:24
Last Modified:	23 Jan 2026 18:12
DOI:	10.24434/j.scoms.2024.01.4270
Related Websites:	Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3180866
Disclaimer:	The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.