Serviços Personalizados
Journal
Artigo
Indicadores
Citado por SciELO
Acessos
Links relacionados
Similares em
SciELO
Compartilhar
PAAKAT: revista de tecnología y sociedad
versão On-line ISSN 2007-3607
Resumo
ARGUELLES TOACHE, Eugenio. Benefits and risks of using Artificial Intelligence in the Mexican Tax Administration Service (SAT). An analysis from the perspective of academic researchers. PAAKAT: rev. tecnol. soc. [online]. 2024, vol.14, n.27, e885. Epub 22-Out-2024. ISSN 2007-3607. https://doi.org/10.32870/pk.a14n27.885.
In the 2024 Master Plan, the Mexican Tax Administration Service (SAT) formally announced that Artificial Intelligence (AI) will be used to classify taxpayers according to their tax risk, identify complex networks of avoidance and evasion, and detect inconsistencies associated with smuggling and front companies. The objective of this work is to identify, analyze and compare the perception of academic researchers on the possible benefits and potential risks of using AI in the SAT. To do so, a questionnaire was developed that was answered by 65 researchers attached to research groups or institutes specialized in AI or tax administration from different Mexican universities. According to these researchers, the possible benefits of using AI in the SAT are optimization of time and resources; greater efficiency and effectiveness in processes; reduction of fraud and tax evasion; greater precision in calculations; reduction of operating costs; increase in tax collection; and improvement of service to taxpayers. On the other hand, the potential risks of using AI in the SAT are the use of algorithms with socioeconomic, racial, national and gender biases that result in processes of discrimination, exclusion and injustice; the disappearance of jobs; processes that become a “black box”; the diffusion of the responsibility of public servants; transparency problems; the generation of unemployment, and unauthorized use of taxpayer data.
Palavras-chave : e-government; smart public administration; tax system; algorithms; machine learning.












