Brazilian Electoral Court Signs Agreement With OpenAI, Anthropic, and ElevenLabs Against Disinformation in the 2026 Elections

Formalized on July 16 in Brasília, the pact includes 7 digital platforms and 3 AI companies — marking the first time model developers have joined the Brazilian electoral program

PrezenceAI Editorial · Published on 20/07/2026

🤖

This article was produced by PrezenceAI with the assistance of artificial intelligence. The main information was extracted from sources cited at the end of this text. The analysis, synthesis, and writing are the sole editorial responsibility of PrezenceAI.

🎧 Listen to this article

0:00 / --:--

What was signed

The Superior Electoral Court (TSE) formalized on July 16, 2026, in a ceremony in Brasília led by Justice Kassio Nunes Marques, cooperation agreements with major digital platforms and — for the first time — with artificial intelligence model developers. OpenAI, Anthropic, and ElevenLabs signed terms of adherence to the Electoral Justice's Permanent Program Countering Disinformation.

Seven digital platforms renewed or signed memorandums of understanding with the TSE: Kwai, Telegram, Meta, TikTok, Google, X, and LinkedIn. The three AI companies joined through a distinct model: Terms of Adherence to the permanent program, a format that recognizes the differing nature of these companies compared to social media networks. Microsoft is expected to formalize its adherence in the coming days.

The focus: deepfakes and generative AI

The agreement specifically expands the fight against the illegal use of artificial intelligence to manipulate candidates' voices and images — so-called deepfakes. The inclusion of ElevenLabs is directly tied to this: the company is a global leader in AI voice synthesis, and its technology was identified as a potential vector for electoral disinformation. Participating companies commit to developing tools and mechanisms to identify new abuse patterns on their platforms. The TSE provides legal backing for moderation and content removal actions.

Context: the program since 2022

The Permanent Program Countering Disinformation was created during the 2022 presidential elections, focusing on narratives that attacked the integrity of electronic voting machines. In March 2026, the TSE approved specific rules on the use of AI during elections — regulations binding on candidates and political parties that prohibit AI providers from allowing uses that violate electoral law. The July agreement expands this cooperation network to include the model developers themselves.

What is still missing

The official statement released by the TSE did not include the full text of the signed documents, nor did it outline specific response deadlines, emergency channels, audit mechanisms, or consequences for non-compliance. The absence of operational details is a limitation highlighted by analysts: intent agreements without measurable SLAs historically suffer from low effectiveness. The October 2026 general elections will serve as the first real test of the agreement's practical reach.