Brazil: Superior Electoral Court (TSE) Regulations

1. Overview

The Supreme Electoral Court (TSE) of Brazil in 2024 issued a resolution No. 23.732/2024  to regulate the use of deepfakes and artificial intelligence (AI) in elections. By preventing voters from being tricked by manipulated content, the law will maintain the integrity of elections.

2. The Scope of the Law

The production and dissemination of artificial intelligence (AI) propaganda, specifically deepfakes, for political purposes is under the law. It prohibits the use of artificial media that manipulates sound or images in order to misrepresent individuals, especially when doing so in an attempt to injure or benefit a candidate. The makers and disseminators of the content are under the law. Content that is obviously satirical or artistic is excluded, as long as there is no intentional attempt at misleading.

3. Key Provisions

• Prohibited Acts: Use of synthetic content in the form of audio, video, or both, that has been digitally produced or altered, even with authorization, to make, replace, or alter the voice or image of an existing, deceased, or imaginary person (deepfake) in order to harm or benefit a candidature is not allowed.
• Conditions of Consent: “Use of any form of multimedia material produced by artificial intelligence in electoral propaganda.” The title’s task is to clearly, prominently, and simply declare that it addresses fabricated or manipulated material and the technology involved.
• Platform Responsibility: On being alerted, online platforms are required to act quickly to delete misinformation. A memorandum of understanding between Meta and other platforms like TikTok and Google ensures that content is deleted within two hours of notification.

4. Enforcement & Penalties

Penalties: Severe penalties for misconduct involve revoking a candidate’s mandate or registration.

Enforcement Authority: Enforcement is left to the TSE in coordination with the Federal Police and the Public Prosecution Service. Sites violating content take-down orders can be subject to legal action.

5. Prominent Precedents or Cases

There were reports of AI-generated content in the 2024 municipal elections. A candidate’s electoral program was halted in one instance where it used AI-generated music without adequate disclosure.

6. Comparison with Global standards

With blanket bans on deepfakes in elections, Brazil has a proactive approach. Whereas the EU AI Act classifies such AI as “high-risk,” Brazil’s legislation directly imposes fines on applicants, reflecting a tough stance.

7. Practical Implications

Candidates must clearly label any content that is generated artificially and remain transparent while using AI during their campaigns. The TSE’s Electoral Disinformation Alert System enables citizens to report content that they consider suspicious. Platforms are required to maintain the collective endeavor towards electoral integrity by responding to takedown notices in a timely manner.

8. Future Prospects

Brazil can consider further amendments as AI technology advances to help respond to new challenges with the manipulation of digital content. Discussions now centre on making regulatory frameworks stronger in protecting democratic processes from advanced disinformation strategies.