ChatGPT as a journalistic writer: Understanding the risks of AI-generated text

Artificial Intelligence, particularly language models like ChatGPT, is revolutionizing text generation. This study investigates the risks of integrating AI-generated content in journalism. A pipeline scrapes articles from seven newspaper websites, aggregating and summarizing data. Analysis of AI-generated summaries reveals two key risks: subtle biases that challenge journalistic integrity and objectivity, and the absence of clear informational hierarchy. The research examines content nuances, style variations, and potential biases in AI-generated text. Findings provide critical insights into the challenges of using machine-generated content in journalism, contributing to a deeper understanding of AI's implications in creative text generation and offering guidance for responsible integration of these technologies in journalistic practices.