During the period from 2024 to 2025, the amount of AI-generated fake content in Brazil increased by 308%.

According to the Brazilian website "opovo," this data comes from a groundbreaking research report titled "Panorama da Desinformação no Brasil" (Panorama of Disinformation in Brazil) released by Lupa Observatory. The study analyzed 617 pieces of internet content verified in 2025 and compared them with 839 pieces verified in 2024, aiming to explore the trends, dissemination purposes, and popular themes of AI-generated fake content.

Research data reveals that in 2024, Lupa Observatory identified 39 pieces of deepfakes and other types of AI-generated fake content in the verified internet content, accounting for 4.6% of the total. In 2025, they found 159 pieces of AI-generated fake content, representing 25% of the total verified content for that year. The number of AI-generated fake content discovered in 2025 increased by 120 pieces compared to the previous year, marking a 308% growth.

Deepfakes refer to highly realistic alterations and synthesis of images, audio, or video using artificial intelligence technology to create content containing false information.

The report highlights that in 2024, AI was mainly used for digital fraud, such as manipulating the voices or images of well-known personalities to promote fraudulent websites. By 2025, this technology was strategically used as a political weapon, with nearly 45% of AI-generated fake content showing clear ideological tendencies, a significant increase from 2024 (33%).

The Lupa Observatory also found that in the AI-generated fake content spread online in 2025, over three-quarters falsified the images or voices of prominent political figures, with 36 pieces related to President Lula, 33 related to former President Bolsonaro, and 30 related to Federal Supreme Court (STF) Justice Alexandre de Moraes.

Data indicates that the proportion of fake content spread through the instant messaging app WhatsApp decreased from nearly 90% in 2024 to 46% in 2025. Analysis suggests that this does not signify a reduction in fake content on WhatsApp but rather indicates a more dispersed distribution of fake content across major communication platforms.

In addition to social networking sites like Facebook, Instagram, Threads, and X, as well as the instant messaging app WhatsApp, short video platforms also witness the dissemination of a significant amount of AI-generated fake content.