Reportagem Investigativa sobre os Sistemas de Decisões Automatizadas

• Autor(a) principal: Carreira, Krishma Anaísa Coura
• Data de Publicação: 2021
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da METODISTA
• Texto Completo: http://tede.metodista.br/jspui/handle/tede/2190
• topic: Sistemas de Decisões Automatizadas; Reportagens investigativas sobre Sistemas de Decisões Automatizadas; Jornalismo Investigativo e Algoritmos; Investigação algorítmica; Impactos dos sistemas automatizados; Automated Decision-Making Systems; Investigative reports on Automated Decision-Making Systems; Investigative Journalism and Algorithms; Algorithmic investigation; Impacts of automated systems; CIENCIAS SOCIAIS APLICADAS::COMUNICACAO

Automated Decision-Making Systems (ADS) have an increasing role in modernity, with profound impacts in the economy, in politics and in society. They can generate numerous benefits, such as productivity boosts and efficiency, but can also miscalculate, increase risks, stimulate inequality, and cause multiple forms of discrimination. Despite this power, the operation of these systems is opaque and at times so complex that it is hard to tell how a decision was taken, or even if it was automated at all. This theme has been the subject of various works in different fields, with discussions ranging from ethics to accountability. In recent times, regulation efforts regarding Artificial Intelligence have also taken place. From the press point of view, ADS tend to become news only when applauded in innovation and tech editorials, or in articles exposing risks, but always lacking the necessary thoroughness. Under the light of Gislene Silva´s (2005) typology, this thesis defends that ADS attend to noticeability criteria based on origin (impact, conflict, controversy, rarity, surprise, government, tragedy/drama, justice and error) and in the vision of facts, especially the public interest noticeability criteria that, in democratic tradition, sparks the need to increase surveillance and critical monitoring of socio-technical systems that involve automated decisions. Thereby, it proposes that they can (and must) be targets of investigative reports, especially when systems are used in public administration. This paper presents a method, named IRADS (Investigative Reports on Automated Decision-Making Systems) that seeks to find ways to identify irregularities, biases, and errors that impact citizens’ lives. The insights here proposed come from an exploratory study, based on a wide revision of interdisciplinary literature by scholars in Communications, Journalism, Sociology, Anthropology, Law, Philosophy of Technology, Tech and Artificial Intelligence. (AU)