Integridade acadêmica na era do ChatGPT

Desafios éticos e as novas fronteiras da Inovação

Abstract

This article investigates the ethical implications of using generative artificial intelligence (GenAI), such as ChatGPT, in the production of scientific research. The analysis is organized into five main categories: information reliability, informational bias, copyright and plagiarism, data security, and inequality of access. The lack of reliability in data is one of the main reasons why AI chatbots should not replace the judgment and responsibility of human researchers. ChatGPT can adopt biases and perpetuate prejudices, in addition to exacerbating social inequalities if its access is not ensured in an equitable manner. AI is widely seen as a tool and cannot be attributed authorship of content, and the absence of proper acknowledgment of original authors may result in plagiarism. Moreover, ChatGPT’s architecture does not fully guarantee the protection of private data collected. The study concludes that the implementation of solid legal frameworks is necessary, requiring the secure and transparent management of information by platforms, alongside the formulation of ethical guidelines for the use of generative AIs by academic institutions and a comprehensive understanding of these tools’ limitations by authors, in order to ensure more responsible research practices.