Recently, an analysis was conducted on the use of artificial intelligence (AI) in opinion columns at Folha, following a notable incident concerning AI utilization in their articles. This study involved the collaboration of AI itself for data extraction and compilation, with the Claude tool from Anthropic providing the coding assistance and Pangram facilitating the evaluation of AI-indicated text. The data collection focused on a total of 3,732 opinion columns published between September 1, 2025, and February 22, 2026, particularly looking at the six months leading up to the controversy surrounding the AI's use in journalism.

The findings revealed that a total of 332 columns, or approximately 8.9%, of the analyzed content included AI-generated text. The extent of AI involvement varied widely, from minor excerpts to complete articles, with a focus on those featuring predominantly AI-produced content. Out of the total analyzed columns, 98 articles contained over 80% AI-written text, which accounted for about 2.6% of the overall evaluated material. Notably, columnist Natalia Beauty had a significant presence in this bracket, with 18 out of 25 published articles showing a majority of AI-generated content.

The implications of these findings raise important questions about the role of AI in editorial processes and the authenticity of human expression in journalism. As AI technology becomes increasingly integrated into the writing landscape, the media may need to navigate ethical considerations concerning transparency and the preservation of individual voice within opinion pieces. The low percentage of AI-generated content currently suggests that while it is present, it remains a relatively minor component of the overall journalistic output at Folha, thus reflecting the ongoing balance between technology and traditional human journalism in Brazil.