A publication made on the official ChatGPT profile generated significant repercussion among Brazilian citizens because the content, presented in English, instructed on how to distinguish between left and right political spectrums. Given the proximity of the second round of the election, the timing of the post ignited intense debates in the comments.
In an exclusive response to Tecnoblog, the artificial intelligence company stated for the first time that there is no connection between this publication and the political campaign held in Brazil.
According to OpenAI, the post was broadcast on a global ChatGPT account with the objective of demonstrating one of the platform's most recent functionalities and illustrating how answers can be presented with greater visual support.
Subsequently, OpenAI formally clarified that such material has no link to the elections or any political campaigns taking place in Brazilian territory.
On Saturday, several Brazilians reacted to the post, using the expression 'fez o L,' a reference to the gesture adopted by supporters of President Luiz Inácio Lula da Silva, who seeks re-election against Flávio Bolsonaro.
Furthermore, major technology companies have become frequent participants in electoral news. Recently, Meta was criticized for allowing the creation of images with an electoral bias directed at candidate Flávio Bolsonaro, but not at candidate Lula.
Tecnoblog investigated that this situation is being treated internally by Meta as an error in the settings of the AI limits, and not as an intentional action. Meta justified that its restriction follows Brazilian electoral norms and, after the incident gained visibility, announced that it implemented complementary actions to prevent commands that used only the candidates' first names.