Detection of texts generated by ChatGPT

capabilities, limitations, and practical applications

Authors

DOI:

https://doi.org/10.35699/2525-8036.2024.53509

Keywords:

Artificial intelligence, ChatGPT, AI detection, Digital security

Abstract

The popularization of Artificial Intelligence (AI) tools for text generation has intensified debates about the authenticity and reliability of digital content amid growing concerns about academic fraud, copyright violations, and misinformation. This study evaluated the effectiveness of five detection tools (Copyleaks AI Content Detector, AI by Smodin, PlagiarismDeTector.NET AI, ZeroGPT, and AI Content Detector) in distinguishing between human-written texts and texts generated by ChatGPT. The results indicate that the tools vary in their accuracy and reliability. While Copyleaks and AI by Smodin demonstrated high accuracy in identifying AI-generated and human-written texts, PlagiarismDeTector.NET AI showed several inconsistencies, especially in correctly distinguishing AI-generated texts. ZeroGPT and AI Content Detector showed variable performance, excelling in some scenarios and facing challenges in others. These findings highlight the ongoing need for improvement in AI text detection tools to ensure a secure and trustworthy digital environment.

Author Biography

  • Felipe Soares de Moraes, Universidade de Brasília

    Mestrando do Programa de Pós-Graduação em Comunicação da Universidade de Brasília (UnB), Brasil. Graduado em Humanidades e em Jornalismo pela Universidade Federal do Sul da Bahia (UFSB). Pesquisador do Centro de Estudos em Comunicação, Política e Tecnologia (CTPol/UnB). ORCID: https://orcid.org/0000-0002-6585-9073. Contato: felipe.moraes2018@gmail.com.  

References

AREMU, Toluwani. Unlocking Pandora's Box: Unveiling the Elusive Realm of AI Text Detection. SSRN, Rochester, NY, 2023. Disponível em: http://dx.doi.org/10.2139/ssrn.4470719. Acesso em: 23 dez. 2024.

CHAKA, Chaka. Detecting AI content in responses generated by ChatGPT, YouChat, and Chatsonic: The case of five AI content detection tools. Journal of Applied Learning and Teaching, v. 6, n. 2, 2023. Disponível em: https://doi.org/10.37074/jalt.2023.6.2.12. Acesso em: 23 dez. 2024.

COUTINHO, Dimítria. 2023, o ano em que o ChatGPT mostrou a IA generativa ao mundo. Data Privacy Brasil, 29 dez. 2023. Disponível em: https://www.dataprivacybr.org/documentos/2023-o-ano-em-que-o-chatgpt-mostrou-a-ia-generativa-ao-mundo/. Acesso em: 23 dez. 2024.

GIL, Antonio Carlos. Métodos e técnicas de pesquisa social. São Paulo: Editora Atlas, 2008.

HUANG, Yue; SUN, Lichao. Harnessing the power of chatgpt in fake news: An in-depth exploration in generation, detection and explanation. Pré-print. arXiv preprint, v. 1, 2023. Disponível em: https://doi.org/10.48550/arXiv.2310.05046. Acesso em: 23 dez. 2024.

LUCCHI, Nicola. ChatGPT: a case study on copyright challenges for generative artificial intelligence systems. European Journal of Risk Regulation, v. 15, n. 3, p. 602-624, 2024. Disponível em: https://www.cambridge.org/core/journals/european-journal-of-risk-regulation/article/chatgpt-a-case-study-on-copyright-challenges-for-generative-artificial-intelligence-systems/CEDCE34DED599CC4EB201289BB161965. Acesso em: 23 dez. 2024.

MEYER, Jesse G.; URBANOWICZ, Ryan J.; MARTIN, Patrick C. N.; O’CONNOR, Karen; LI, Ruowang; PENG, Pei-Chen; BRIGHT, Tiffani J.; TATONETTI, Nicholas; WON, Kyoung Jae; GONZALEZ-HERNANDEZ, Graciela; MOORE, Jason H. ChatGPT and large language models in academia: opportunities and challenges. BioData Mining, v. 16, n. 1, 2023. Disponível em: https://doi.org/10.1186/s13040-023-00339-9. Acesso em: 23 dez. 2024.

OPENAI. OpenAI Charter. São Francisco: 2024. Disponível em: https://openai.com/charter/. Acesso em: 23 dez. 2024.

SARKAR, Ratan Kumar. AI Content Detector. ChatGPT, 2024. Disponível em: https://chatgpt.com/g/g-NcgzdHmZc-ai-content-detector. Acesso em: 23 dez. 2024.

SARRION, Eric. Exploring the Power of ChatGPT: Applications, Techniques, and Implications. Berkeley: Apress, 2023.

TRINDADE, Alessandra Stefane C. E.; OLIVEIRA, Henry Poncio C. Inteligência artificial (IA) generativa e competência em informação: habilidades informacionais necessárias ao uso de ferramentas de IA generativa em demandas informacionais de natureza acadêmica-científica. Perspectivas em Ciência da Informação, v. 29, p. e-47485, 2024. Disponível em: https://doi.org/10.1590/1981-5344/47485. Acesso em: 23 dez. 2024.

Published

2024-12-24

How to Cite

Detection of texts generated by ChatGPT: capabilities, limitations, and practical applications. Revista de Ciências do Estado, Belo Horizonte, v. 9, n. 2, p. 1–21, 2024. DOI: 10.35699/2525-8036.2024.53509. Disponível em: https://periodicos.ufmg.br/index.php/revice/article/view/e53509. Acesso em: 25 dec. 2024.

Similar Articles

1-10 of 54

You may also start an advanced similarity search for this article.