Publicação
AI-based consumer's creditworthiness assessment: era of automation, future of scoring and the EU policymaking on automated decision-making
| Resumo: | Today, credit data drives almost the entire consumer lending operation. Applicants should fear how some of their demographic, financial, employment or behavioural characteristics affect (may affect) determinately the possibility of obtaining loans. Credit scoring, fundamentally, stands as a tool that lies its value at the pre-contractual stage of determining the passive party. It is no longer the credit analysts or the programmers but the inputs sets' quality and, hence, the self-learning models derived, that decides whom to be granted a loan. From traditional judgemental systems to recent technological breakthroughs, AI software have shown an increasingly ability to operate successfully in classification tasks such as creditworthiness assessment. However, scoring based on AI raises an energetic tutelage on protecting personal data, especially in what esteems profiling consumers’ solvency. Are the GDPR and the EU sectorial policymaking ready to meet the challenges exhorted by Big Data and AI? How lawful is it for lenders and bureau agencies to rely on alternative data to assess a client’s creditworthiness? How or when credit analysts must intervene? What kind of information should they provide to the data subjects? Thus, it was in the light of the scope, legal grounds, and automated decision-making regime, as well as the somewhat illusory guarantees that the European legislator has enshrined - in Art. 22 of the GDPR, Arts. 13 and 14 of the Proposal for an AI Act, and in the Arts. 12 and 18 (6)(a)(b)(c) of the Proposal for a Directive on Consumer Credits, of 30 June 2021 - that we conclude the need to adopt multidisciplinary regulatory policies striving for a better (cyber) consumers’ financial info literacy. |
|---|---|
| Autores principais: | Rebelo, Diogo Morgado |
| Outros Autores: | Ferreira, Filipa Campos |
| Assunto: | AI-based scoring Creditworthiness assessment Automated decision-making Consumer Credit Scoring baseado em IA Decisões automatizadas Avaliação de solvabilidade |
| Ano: | 2022 |
| País: | Portugal |
| Tipo de documento: | capítulo de livro |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | Today, credit data drives almost the entire consumer lending operation. Applicants should fear how some of their demographic, financial, employment or behavioural characteristics affect (may affect) determinately the possibility of obtaining loans. Credit scoring, fundamentally, stands as a tool that lies its value at the pre-contractual stage of determining the passive party. It is no longer the credit analysts or the programmers but the inputs sets' quality and, hence, the self-learning models derived, that decides whom to be granted a loan. From traditional judgemental systems to recent technological breakthroughs, AI software have shown an increasingly ability to operate successfully in classification tasks such as creditworthiness assessment. However, scoring based on AI raises an energetic tutelage on protecting personal data, especially in what esteems profiling consumers’ solvency. Are the GDPR and the EU sectorial policymaking ready to meet the challenges exhorted by Big Data and AI? How lawful is it for lenders and bureau agencies to rely on alternative data to assess a client’s creditworthiness? How or when credit analysts must intervene? What kind of information should they provide to the data subjects? Thus, it was in the light of the scope, legal grounds, and automated decision-making regime, as well as the somewhat illusory guarantees that the European legislator has enshrined - in Art. 22 of the GDPR, Arts. 13 and 14 of the Proposal for an AI Act, and in the Arts. 12 and 18 (6)(a)(b)(c) of the Proposal for a Directive on Consumer Credits, of 30 June 2021 - that we conclude the need to adopt multidisciplinary regulatory policies striving for a better (cyber) consumers’ financial info literacy. |
|---|