Brazil's Superior Electoral Court is implementing tighter rules to combat the influence of artificial intelligence in the 2026 general elections.
The rise of ultra-realistic AI content poses a significant risk to democratic integrity by enabling the spread of misinformation through deep-fakes. Because these tools can manipulate voter perception on a national scale, the court is under pressure to establish effective safeguards before the campaign begins.
Kássio Nunes Marques assumed the presidency of the Tribunal Superior Eleitoral (TSE) on May 12, 2026 [1]. Under his leadership, the court is navigating the tension between technological innovation and electoral security. The primary concern involves AI-generated audio, images, and videos that can be weaponized to deceive the public [2, 3].
To address these threats, Supreme Court minister Gilmar Mendes advocated for the creation of a specialized task force on July 3, 2026 [2]. This proposed group would focus specifically on identifying and neutralizing deep-fakes in real time. The move follows growing warnings that AI could become a central tool for political manipulation during the upcoming cycle [4].
However, some experts argue that the current regulatory framework is not enough. Political marketing consultant Marcelo Vitorino said in a July 22, 2026, interview that the TSE's rules are insufficient to address the scale of AI-driven manipulation [3]. This critique comes as the country approaches the official campaign period, with less than one month remaining as of late July [3].
AI expert Giovanni La Porta also highlighted the systemic danger, noting on July 14, 2026, that AI represents a direct threat to democracy [4]. The debate has already divided pre-campaign presidential candidates, who hold differing views on whether AI should be permitted in political advertising [5].
While the TSE seeks a proactive stance, the speed of AI development continues to outpace legislation. The court must now balance the protection of free speech with the need to prevent synthetic media from distorting the electoral outcome.
“AI-generated audio, images, and videos can be weaponized to deceive the public.”
The conflict between the TSE's regulations and the capabilities of generative AI suggests a looming crisis of trust in digital evidence. If the court cannot effectively distinguish between authentic and synthetic media, the 2026 election may be defined not by policy debates, but by a battle over the authenticity of information, potentially undermining the legitimacy of the final results.
