With the evolution of artificial intelligence (AI), more and more content is generated by generative AI and is reproduced in the digital environment. However, common aspects in stories, essays and articles highlight the tool’s language patterns, which can be identified by readers.
Some tips for identifying content created by AI have already been made available by ChatGPT itself, a free program that has become a popular chatbot.
Among the tools that facilitate the possible identification of readers are, for example, linguistic, content and coherence standards.
See below how to identify AI-generated texts
The lack of vocabulary variation and, at the same time, the use of clichéd expressions are some of the linguistic patterns that help to detect content generated by generative AI, in addition to the inappropriate use of slang.
The repetition of generalizations and the lack of development of ideas reveal the standard of content generated by such a tool.
The absence of strategies that can generate different interpretations is also a hallmark of texts generated by AI, such as metaphors and figures of speech.
Furthermore, the repetition of information, the lack of precise conclusions and the constant mention of external references show the pattern of coherence in content generated by generative AI.
Typing and spelling errors, as well as the lack of authorship in the texts, are other factors that make it possible to identify the creation of materials through algorithms and data sets.
This content was originally published in Texts written by AI; Is it possible to identify? on the CNN Brasil website.
Source: CNN Brasil

Charles Grill is a tech-savvy writer with over 3 years of experience in the field. He writes on a variety of technology-related topics and has a strong focus on the latest advancements in the industry. He is connected with several online news websites and is currently contributing to a technology-focused platform.