Jornalismo na era da IA: limites e possibilidades da inteligência artificial generativa em veículos jornalísticos

Abstract

This dissertation investigates how textual generative artificial intelligence is used in two journalistic communication products, both operating in the digital environment: the Brazilian Nuclito and the North American Realtime, especially in the creation of news content, in addition to its the impact on users. Market issues, regulation in Brazil and abroad, and the historical trajectory of AI to the present day are also addressed. Regarding methodological strategies, qualitative case studies with comparative analyses are used, such as ones authored by Yin (2005), Lakatos and Marconi (2017), and Prodanov and Freitas (2013), as well as the collective case study by Stake (1995). It is concluded that each product has its own strategy, but both monitor issues such as hallucinations, possible biases, and accuracy in results.