Brazil: Mandatory identification of AI-generated content and objective liability of digital platforms (PL 83/2026) was introduced to Chamber of Deputies

On 2 February 2026, a mandatory identification of content generated or manipulated by artificial intelligence (AI) (PL 83/2026) was introduced to the Chamber of Deputies. The Bill would apply to users and digital platforms that publish, share, recommend, monetise or disseminate synthetic content, including videos, images, audio and text. It would require clear, conspicuous and permanent labelling of AI-generated or substantially manipulated content circulated on digital platforms and social networks. Users would be required to declare the synthetic nature of content at the time of publication. Digital platforms would be held objectively liable for disseminating unlabelled synthetic content if they allow publication without the required labelling, recommend or monetise the content, or fail to act immediately after becoming aware of infringement. The Bill would also prohibit the dissemination of certain deepfakes relating to electoral processes, public authorities, security forces or the constitutional branches of government. It would also establish criminal penalties of 2 to 6 years' imprisonment and fines for deepfakes that could cause collective or institutional harm. It would provide for administrative sanctions, including fines of up to 20 % of the gross turnover of the economic group in Brazil, limited to BRL 100 million per infringement, temporary suspension of platform functionalities, prohibition of monetisation, and content blocking within the national territory. The Bill will now proceed through the legislative process.