Brazil's TSE signs pact with tech giants to curb disinformation

The Tribunal Superior Eleitoral (TSE) signed a memorandum of understanding with major digital platforms – Google, X, Meta, Kwai, Telegram, TikTok and LinkedIn – and three artificial‑intelligence firms, OpenAI, ElevenLabs and Anthropic. The agreement, signed on 16 July 2026, expands the permanent programme that has been in place since the 2022 presidential election to fight false narratives that could threaten the integrity of electronic voting and the legitimacy of the 2026 elections.

The pact intensifies actions against the illegal use of artificial intelligence, prohibiting the creation, publication or amplification of AI‑generated images, voice or video of candidates within the 72‑hour window before voting until 24 hours after polls close. Candidates must clearly label any content produced by AI, and platforms are required to remove deep‑fake or misogynistic material that manipulates candidates’ appearance. The TSE’s own Resolution No. 23.755, issued on 2 March 2026, imposes similar labeling and timing rules for AI content across all campaign media.

Anne Cabral, election‑law specialist, warned that “we cannot demonise artificial intelligence; what we demonise is the deep‑fake.” The first round of the election is scheduled for 4 October, with a possible runoff on 25 October.