O efeito do ChatGPT nas solicitações de patentes: para o caso brasileiro

• Ano de defesa: 2025
• Autor(a) principal: Silva, Carlos Leonardo Fernandes da
• Orientador(a): Paiva, Matheus Silva de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Católica de Brasília
• Programa de Pós-Graduação: Programa Stricto Sensu em Governança, Tecnologia e Inovação
• Departamento: Escola de Educação, Tecnologia e Comunicação
• País: Brasil
• Palavras-chave em Português: ChatGPT; Patentes; Inteligência artificial
• Palavras-chave em Inglês: Artificial intelligence; Patents
• Área do conhecimento CNPq: CNPQ::CIENCIAS SOCIAIS APLICADAS::CIENCIA DA INFORMACAO
• Link de acesso: https://bdtd.ucb.br:8443/jspui/handle/tede/3692

This dissertation investigates the effect of the ChatGPT introduction on the volume of patent applications in Brazil, within a broader context of socioeconomic and institutional determinants of formal innovation. Acknowledging the growing relevance of generative artificial intelligence in intellectual property, the study analyzes how ChatGPT, as a natural language processing and generation tool, could influence the patenting process in the country. A quantitative and explanatory research was conducted, employing multiple linear regression based on time series data from 2000 to 2023. The econometric model considered the annual number of patent deposits as the dependent variable and included independent variables such as GDP per capita, SELIC rate, government investment in R&D, CNPq scholarships, total public spending, average schooling, the COVID-19 pandemic, and the introduction of ChatGPT (the latter two as dummy variables). The regression results indicated that GDP per capita, government investment in R&D, and total public spending had positive and statistically significant impacts on patent applications, reinforcing the role of economic development and public policies in stimulating formal innovation. The COVID-19 pandemic, in turn, showed a negative and significant effect on patent deposits. However, the variable representing the introduction of ChatGPT did not present statistical significance during the analyzed period. Diagnostic tests pointed to the presence of autocorrelation in the model residuals, suggesting the need for methodological improvements in future studies. Despite the limitations, the research contributes to the understanding of the dynamics of patented innovation in Brazil and raises important questions about measuring the impact of emerging technologies like ChatGPT, which may influence innovation in ways not yet fully captured by traditional intellectual property indicators. It is concluded that, although the direct effect of ChatGPT on patent applications was not statistically proven in the studied period, the increasing importance of AI for innovation demands careful attention and the development of methodologies that capture its complex and multifaceted impacts.