Os sistemas automatizados para análise de crédito pelas instituições financeiras e os abusos ao direito do consumidor
Tipo
TCC
Data de publicação
2023-12
Periódico
Citações (Scopus)
Autores
Luca, Natajsha Simonsen de
Orientador
Padin, Camila Ferrara
Título da Revista
ISSN da Revista
Título de Volume
Membros da banca
Programa
Resumo
A dissertação tem por objetivo principal analisar as possíveis abusividades no direito do
consumidor a partir de decisões automatizadas de análise de crédito pelas instituições
financeiras. A Inteligência Artificial surgiu para auferir agilidade e eficiência, no entanto, esta
tecnologia nem sempre é neutra, exatamente em razão de que esta é programada por humanos,
que contém preconceitos. A utilização de Inteligência Artificial para pontuação de crédito, com
algoritmos que funcionam de forma enviesada, tem o potencial de ensejar julgamentos
arbitrários e a discriminação de grupos historicamente marginalizados e menos privilegiados
em razão do uso inadequado de dados ou características associadas a eles. Os programadores
destas instituições criaram algoritmos preditivos a consumidores vulneráveis, que não avaliam
tão somente as características de risco do cliente, mas analisam o risco do ambiente em que um
consumidor está utilizando crédito, o que pode contribuir com a perpetuação dos ciclos de
pobreza. Apesar de concederem empréstimos a estes clientes que possuem históricos de crédito
limitados, muitas vezes exigem taxas de juros mais altas. A ausência de acesso ao crédito em
condições equitativas e razoáveis pode acarretar custos exorbitantes para indivíduos de baixa
renda, ao passo que impossibilita o atendimento de suas obrigações financeiras de curto prazo.
Assim, decisões automatizadas que carecem de ética e prudência têm o potencial de resultar em
diversos prejuízos para o consumidor, o qual já é considerado a parte mais vulnerável na relação
de consumo, tornando-se ainda mais exposto e suscetível às manipulações diante das novas
tecnologias. Este cenário agrava-se com a falta de transparência das decisões automatizadas, o
que aumenta a assimetria informacional entre consumidor e instituição financeira, tornando-se
imprescindível a regulamentação específica da inteligência artificial para a proteção dos direitos
e da privacidade dos cidadãos, fomentando a transparência e a explicabilidade dos sistemas
providos de inteligência artificial e mitigando potenciais riscos. Para tanto, utiliza-se a pesquisa
bibliográfica-documental com abordagem de natureza qualitativa e método hipotético-dedutivo.
The main objective of this dissertation is to analyze the possible abuse of consumer rights as a result of automated credit analysis decisions by financial institutions. Artificial Intelligence has emerged to bring agility and efficiency; however, this technology is not always neutral, precisely because it is programmed by humans, who contain prejudices. The use of Artificial Intelligence for credit scoring, with algorithms that work in a biased way, has the potential to lead to arbitrary judgments and discrimination against historically marginalized and less privileged groups, due to the inappropriate use of data or characteristics associated with them. The programmers at these institutions have created predictive algorithms for vulnerable consumers, which not only assess the risk characteristics of the client, but also analyze the risk of the environment in which a consumer is using credit, which can contribute to perpetuating cycles of poverty. Despite granting loans to these clients who have limited credit histories, they often demand higher interest rates. The lack of access to credit on fair and reasonable terms can lead to exorbitant costs for low-income individuals, while making it impossible for them to meet their short-term financial obligations. Thus, automated decisions that lack ethics and prudence have the potential to result in various losses for the consumer, who is already considered the most vulnerable party in the consumer relationship, becoming even more exposed and susceptible to manipulation in the face of new technologies. This scenario is exacerbated by the lack of transparency in automated decisions, which increases the information asymmetry between the consumer and the financial institution, making it essential to specifically regulate artificial intelligence in order to protect citizens' rights and privacy, fostering transparency and explainability in artificial intelligence systems, and mitigating potential risks. To this end, bibliographical-documentary research is used, with a qualitative approach and a hypothetical-deductive method.
The main objective of this dissertation is to analyze the possible abuse of consumer rights as a result of automated credit analysis decisions by financial institutions. Artificial Intelligence has emerged to bring agility and efficiency; however, this technology is not always neutral, precisely because it is programmed by humans, who contain prejudices. The use of Artificial Intelligence for credit scoring, with algorithms that work in a biased way, has the potential to lead to arbitrary judgments and discrimination against historically marginalized and less privileged groups, due to the inappropriate use of data or characteristics associated with them. The programmers at these institutions have created predictive algorithms for vulnerable consumers, which not only assess the risk characteristics of the client, but also analyze the risk of the environment in which a consumer is using credit, which can contribute to perpetuating cycles of poverty. Despite granting loans to these clients who have limited credit histories, they often demand higher interest rates. The lack of access to credit on fair and reasonable terms can lead to exorbitant costs for low-income individuals, while making it impossible for them to meet their short-term financial obligations. Thus, automated decisions that lack ethics and prudence have the potential to result in various losses for the consumer, who is already considered the most vulnerable party in the consumer relationship, becoming even more exposed and susceptible to manipulation in the face of new technologies. This scenario is exacerbated by the lack of transparency in automated decisions, which increases the information asymmetry between the consumer and the financial institution, making it essential to specifically regulate artificial intelligence in order to protect citizens' rights and privacy, fostering transparency and explainability in artificial intelligence systems, and mitigating potential risks. To this end, bibliographical-documentary research is used, with a qualitative approach and a hypothetical-deductive method.
Descrição
Palavras-chave
inteligência artificial , instituição financeira , análise de crédito , decisões autônomas , artificial intelligence , financial institution , credit analysis , autonomous decisions