Region · AI content creation

AI content creation in Brazil.

Brazil is a large, creator-led market where AI-made content works if it sounds Brazilian rather than translated. The rules that apply now are the LGPD for faces and voices, the Consumer Defence Code and CONAR for advertising, and, since March 2026, a ban on profiling-based ads aimed at children and teenagers. The AI bill passed by the Senate in December 2024, which would require synthetic content to be identified, is still before the Chamber of Deputies.

LGPD, image rights and the data agency

The General Data Protection Law (Law 13.709/2018) treats biometric data as sensitive, so building a digital twin or voice model needs specific, highlighted consent or another narrow legal basis. The regulator became a full regulatory agency in September 2025, confirmed by law in February 2026, keeping the ANPD acronym and gaining a remit that includes children's digital protection. Image rights are strong. The Constitution protects a person's image, the Civil Code lets people stop commercial use of it, and the Superior Court of Justice has held that unauthorised commercial use of someone's image is compensable without proof of harm. A lookalike, soundalike or synthetic version of a real person carries real exposure. The heirs of a deceased person can also protect that person's image.

CONAR, the Consumer Defence Code and one famous case

The Consumer Defence Code (Law 8.078/1990) requires advertising to be easy to identify as advertising and bans misleading and abusive ads, including those that exploit children's inexperience. CONAR, the self-regulatory council, applies its own code and guidance on influencer advertising. In 2023 it examined a car advert that used AI to put a late singer beside her daughter. The case was archived: the heirs had consented, and a majority found the AI use obvious, with no specific rule at the time. It still made consent from estates and AI disclosure part of CONAR's agenda. Since 17 March 2026, the ECA Digital (Law 15.211/2025) has banned using profiling to target commercial advertising at children and adolescents.

The AI bill and the direction of travel

Bill PL 2338/2023, Brazil's proposed AI framework, passed the Senate on 10 December 2024 and sits with a special committee of the Chamber of Deputies, which had not voted on it by September 2026. The Senate text requires synthetic content generated by AI to carry an identifier of its origin, with lighter options such as credits or metadata for artistic works, and it includes rules on copyright in training data. If the Chamber changes the text, it returns to the Senate. The electoral court already requires labels on AI content in campaign advertising and bans deepfakes of candidates. For brands, the sensible course is to build identification into realistic synthetic content now rather than retrofit it.

Brazilian Portuguese, WhatsApp and creators

Brazilian Portuguese differs from European Portuguese in vocabulary, grammar, pronouns and pronunciation, and Brazilians notice imported copy instantly. Tone is warmer and more informal than in most European markets, and regional accents from São Paulo, Rio de Janeiro or the Northeast say different things about a brand. Commerce runs through WhatsApp, where customers ask questions, receive offers and pay, often by Pix instant transfer. Creators drive discovery on Instagram, TikTok, YouTube and Kwai, and Mercado Livre and Shopee lead online retail. Generated content that looks like a polished global ad can underperform creator-style content here, so brief for the formats and voices that Brazilians already follow, and disclose the brand behind every piece.

Updated 25 September 2026 · General information, not legal advice. Rules change, so check the current text with your legal team before relying on it.

Questions

AI content creation in Brazil, answered.

Do AI-generated ads need a label in Brazil?

There is no general statutory AI label yet. Ads must be identifiable as advertising and must not mislead, and CONAR expects honesty about content that could deceive. The pending AI bill would require identifiers on synthetic content, and electoral rules already require AI labels in campaign ads, so labelling realistic synthetic content is the prudent default.

Can we recreate a deceased celebrity with AI in Brazil?

Only with the consent of the heirs, who can protect the person's image. CONAR's 2023 case on a car advert featuring a late singer was archived because the family had consented and the AI use was obvious. Beyond the law, audiences judge these ads on respect, so involve the family in the creative, not only the contract.

Can we reuse European Portuguese content in Brazil?

Not without adaptation. Vocabulary, grammar, pronouns and spelling differ, from ônibus against autocarro to the way people say 'you', and a European Portuguese voice sounds foreign in Brazil. Keep Brazilian and European Portuguese as separate versions, each with its own glossary, voices and native reviewer.

Get started

See your AI studio generating this week.

Thirty minutes. A live studio in your colours, your brand hub loaded, a real campaign brief.

See plans