AI Regulation Tracker / Bill approved by Senate

Brazil Senate Approves PL 3066/2025 Criminalizing AI and Deepfake Child Abuse Imagery

Regulatory summary: Brazil's Senado Federal approved PL 3066/2025 on July 7, 2026, extending child sexual-abuse-material crimes to content produced or generated by AI and deepfakes even when no real person is depicted, raising penalties and adding an AI aggravator. It now goes to presidential sanction.

The bill treats synthetic and AI-generated child sexual abuse material as a crime even when no real child is depicted, and now awaits presidential sanction.

Primary source

By Anthony Guerriero, Founder, The Leveraged Years · Reviewed by The Leveraged Years Editorial Desk · Published July 9, 2026 · Last updated July 9, 2026

Brazil Senate Approves PL 3066/2025 Criminalizing AI and Deepfake Child Abuse Imagery regulation briefing

The Leveraged Years AI Regulation Tracker

Key takeaways

The Senate extended child sexual-abuse-material offenses to representations produced, manipulated, or generated by digital technologies, including artificial intelligence and deepfakes, even when they do not depict a real person. It raised penalties and added a one-third to two-thirds aggravator when AI, deepfakes, or anonymization is used.
Criminal-defense and family attorneys, platform trust-and-safety teams, compliance officers, DPOs, EdTech and child-services operators, and content-moderation and detection vendors operating in or serving Brazil
Status: Approved by the Senate July 7, 2026.
Before sanction, review detection and reporting policies against a definition that covers fully synthetic material, and prepare legal and trust-and-safety teams for a heinous-crime classification that removes leniency benefits

Regulatory Briefing

Bindingness: Bill approved by Senate, awaiting presidential sanction (not yet enacted)
Scope: Synthetic and AI-generated child sexual abuse material criminalization; increased penalties; AI/deepfake aggravator; heinous-crime classification
What Changed: The Senate extended child sexual-abuse-material offenses to representations produced, manipulated, or generated by digital technologies, including artificial intelligence and deepfakes, even when they do not depict a real person. It raised penalties and added a one-third to two-thirds aggravator when AI, deepfakes, or anonymization is used.
Who This Affects: Criminal-defense and family attorneys, platform trust-and-safety teams, compliance officers, DPOs, EdTech and child-services operators, and content-moderation and detection vendors operating in or serving Brazil
Current Status: Approved by the Senate July 7, 2026; sent to presidential sanction
Effective Date: Not yet in force (pending sanction and publication)
Primary Source: https://www12.senado.leg.br/noticias/materias/2026/07/07/aprovadas-penas-mais-severas-para-violencia-sexual-digital-contra-menores
Secondary Analysis: Agencia Brasil, "Congresso aumenta punicao a crimes sexuais online contra criancas" (July 2026); Agencia Senado plenary coverage
Workflow Impact: Platforms and detection vendors serving Brazilian users would face content that qualifies as criminal even absent a real victim, changing hash-matching, classifier tuning, escalation, and law-enforcement reporting assumptions
Operational Next Step: Before sanction, review detection and reporting policies against a definition that covers fully synthetic material, and prepare legal and trust-and-safety teams for a heinous-crime classification that removes leniency benefits
Last Verified: July 9, 2026

Date	Jurisdiction	Rule	Affected professionals	Status or effective date
2026-07-09	Brazil	The Senate extended child sexual-abuse-material offenses to representations produced, manipulated, or generated by digital technologies, including artificial intelligence and deepfakes, even when they do not depict a real person. It raised penalties and added a one-third to two-thirds aggravator when AI, deepfakes, or anonymization is used.	Criminal-defense and family attorneys, platform trust-and-safety teams, compliance officers, DPOs, EdTech and child-services operators, and content-moderation and detection vendors operating in or serving Brazil	Approved by the Senate July 7, 2026; sent to presidential sanction

Frequently Asked Questions

Is PL 3066/2025 already law in Brazil?

No. The Senado Federal approved the bill on July 7, 2026, and sent it to presidential sanction. It becomes law only after the President sanctions it and the text is officially published. Until then it is an approved bill awaiting sanction, not enacted legislation.

Does the bill criminalize AI-generated abuse imagery that involves no real child?

Yes, under the approved text. It extends the covered crimes to material produced, manipulated, or generated by digital technologies, including artificial intelligence and deepfakes, even when the content does not depict a real person.

What are the new penalty ranges?

According to the Senado Federal, possession offenses move to 3 to 6 years of reclusao plus a fine, and production and distribution offenses move to 4 to 10 years of reclusao plus a fine. A one-third to two-thirds increase applies when AI, deepfakes, or anonymization is used.

What does the hediondo classification mean in practice?

Classifying these offenses as heinous crimes restricts leniency benefits, tightens conditions for provisional release, and lengthens custodial-regime progression requirements. It applies to both consummated and attempted forms under the approved text.

What should trust-and-safety and compliance teams do before sanction?

Review detection and classifier logic so it covers fully synthetic material, confirm that escalation and reporting workflows do not depend on identifying a real victim, and brief legal teams on the penalty increases and heinous-crime classification so they are ready when and if the bill is sanctioned.