Brazil ANPD Monitors AI Intimate-Content Risks
Brazil’s data-protection authority says Decrees 12.975/2026 and 12.976/2026 are in force and has begun monitoring platform duties, including risks from AI-generated or AI-modified intimate content. This is not a Brazilian AI Act, a new ANPD AI regulation, or a published enforcement action.
What is in force: ANPD says Decree 12.975 updated Marco Civil da Internet regulation for internet application providers, including platforms and social networks. It also says Decree 12.976 addresses protection of women in digital settings, including intimate content produced, modified or disseminated through digital technologies, including AI.
What ANPD is doing: Its June to July 2026 timetable includes monitoring provider duties and monitoring generation and modification of intimate content by AI.
What this is not: The ANPD page does not announce an AI Act, a technical detection standard, a fine, a finding against a company, or a new enforcement case.
- Authority
- Autoridade Nacional de Proteção de Dados, Brazil.
- Instruments
- Decree 12.975/2026 and Decree 12.976/2026, both dated 20 May 2026.
- Status
- ANPD states the decrees are in force. Its June to July timetable describes implementation monitoring.
- Primary source
- ANPD Marco Civil da Internet page
What does ANPD say has changed?
ANPD says that Decree 12.975/2026 updated the regulation of Marco Civil da Internet and established new obligations for internet application providers, including digital platforms and social networks. On the page, ANPD identifies proactive measures, transparency, prevention and mitigation of risks, and measures addressing digital fraud and scams among the areas affected. It also refers to unavailability of criminal content in cases provided for by the instrument.
The same ANPD record says Decree 12.976/2026 establishes guidelines for protection of women in the digital environment. The page specifically includes production, modification and dissemination of intimate content through digital technologies, including artificial intelligence. That is the verified AI connection in the source.
The authority’s page is an implementation record. It does not provide a complete decree-by-decree compliance manual. It therefore cannot support claims about every covered service, a technical content-detection method, exact response windows, penalties, defences, or a universal AI disclosure rule.
What is ANPD monitoring?
ANPD’s first implementation stage covers orientation, institutional communication, social listening and initial monitoring. The June to July timetable states that ANPD will monitor compliance with provider duties under the two decrees. The page refers to general duties, duty of care, notification and unavailability of criminal content, advertisements, paid boosts and publicity.
Separately, the timetable identifies monitoring of the generation and modification of intimate content by artificial intelligence. The stated focus is the risk related to production and dissemination of AI-manipulated intimate content. Monitoring does not itself establish that ANPD has found a particular company in breach. The source gives no fine, order, named target, or enforcement finding.
| Question | What ANPD’s page supports |
|---|---|
| Are the two decrees in force? | Yes. ANPD states that Decrees 12.975/2026 and 12.976/2026 entered into force. |
| Is this a general Brazilian AI Act? | No. The page concerns Marco Civil platform regulation and digital protection measures, including an AI-related intimate-content risk. |
| Has ANPD announced a new AI enforcement action? | No. The page describes monitoring and implementation activity, not an enforcement decision. |
| Does the page prescribe a technical AI-detection tool? | No such technical requirement is identified on the ANPD page. |
What should platform teams do with this update?
Platform teams with Brazil-facing services should treat this as an evidence-preservation and governance signal. A useful immediate step is to identify which internal functions own notices, risk assessment, content restrictions, paid-placement review and escalation of reports involving intimate content. That is not a substitute for reading the operative decrees or obtaining Brazilian legal advice. It is a practical way to make later source-specific review possible.
Teams should keep the AI point narrow. The ANPD page says it will monitor the generation and modification of intimate content by AI and related production and dissemination risks. It does not say that every AI-generated image is prohibited, that every platform must use a specific classifier, or that an AI provider has received a new standalone ANPD duty.
Why this deserves tracking
The significance is the combination of in-force platform regulation and a regulator’s published monitoring plan that expressly names AI-manipulated intimate content. That is more concrete than a general policy statement, but it is still not a blank cheque to infer additional requirements. The next authoritative materials to watch are the decree texts, ANPD guidance, a notice, or a specific enforcement action.
ANPD says Brazil’s platform decrees are in force and is monitoring AI-related intimate-content risks. Keep the status exact: this is an in-force platform-regulation update and monitoring program, not a general AI law or a published enforcement case.
Marco Civil da Internet, ANPD. To verify, confirm the statements that Decrees 12.975/2026 and 12.976/2026 entered into force and the June to July monitoring timetable.
FAQ
Has Brazil passed a general AI Act through these decrees?
No. ANPD’s page describes Marco Civil platform-regulation changes and digital-protection guidelines that include AI-manipulated intimate-content risks. It does not describe either decree as a general Brazilian AI Act.
What does ANPD say it is monitoring?
ANPD’s June to July 2026 timetable includes monitoring provider duties under the two decrees and monitoring generation and modification of intimate content by AI, focused on production and dissemination risks.
Has ANPD fined or named a company over AI intimate content?
No. The ANPD page describes implementation monitoring. It does not identify a fine, enforcement order, named target, or non-compliance finding.
Does ANPD require a specific AI-content detection tool?
No. The ANPD page does not identify a specific detection technology, technical standard, or universal AI-content labeling requirement.