Key Points
—Brazil’s antitrust authority CADE voted unanimously on April 23 to escalate its six-year-old investigation into Google’s use of journalistic content from an administrative inquiry to a full administrative process.
—The new proceeding explicitly includes Google’s generative AI features in search results — the first time a Latin American antitrust regulator has placed AI-generated content at the centre of a market-dominance case.
—The case had been shelved by CADE’s general superintendence in December 2024 for insufficient evidence. Interim CADE chief Diogo Thomson de Andrade led the recommendation to revive it, citing “the evolution of Google’s conduct.”
—Google called the CADE decision a “misunderstanding” of how its products work and said it will continue to engage. The Brazilian National Newspaper Association has pushed for mandatory compensation deals with publishers.
Brazil has decided not to let Google close the book on a six-year case. It has also decided that the book now includes AI. Both moves matter for investors watching how Latin America’s largest economy will treat the American companies at the centre of the generative-AI boom.
The Rio Times, the Latin American financial news outlet, reports that CADE, Brazil’s Administrative Council for Economic Defense, voted unanimously on April 23 to escalate its investigation into Alphabet’s Google to a formal administrative process. The Brazil CADE Google AI case had been effectively dormant since December 2024, when the agency’s general superintendence recommended shelving it for lack of sufficient evidence of anti-competitive conduct.
The tribunal’s reversal on Thursday was driven by interim CADE chief Diogo Thomson de Andrade, who returned the case to the general superintendence for formal proceedings. He cited “the evolution of Google’s conduct” since the original 2019 inquiry began.
Why the Brazil CADE Google AI decision is different from the 2019 case
The original 2019 investigation focused on Google’s automated collection of journalistic content and its display within search results — the same complaint that European regulators and Australian legislators have been processing for years. The question was traffic and advertising-revenue capture. The mechanism under scrutiny was snippets, knowledge panels, and news carousels.
The reopened case is different. CADE’s reasoning explicitly cites generative AI features — the AI-generated summaries that appear above traditional search results in Google’s product, and the degree to which those summaries are built on crawled journalistic content without negotiated compensation.
That is a new analytical frame. It asks whether Google’s dominant position in search is now being leveraged into a dominant position in AI-mediated information retrieval, at the expense of the news-content producers on whose work the AI system is trained and grounded.
No other Latin American antitrust authority has placed generative AI at the centre of a market-dominance case. That makes CADE’s Thursday decision a regulatory first for the region.
How a shelved case came back
The case was opened in 2018 and formalised in 2019 after CADE determined it would examine competitive conditions in Brazil’s search and news markets. It ran for five years largely below the headline level until December 2024, when the general superintendence recommended closing it.
In March 2025, CADE Commissioner Camila Cabral Pires Alves went on the record recommending that the case be reopened. She cited mounting regulatory scrutiny of digital platforms in other jurisdictions — specifically the European Commission’s investigation into whether Google had violated the Digital Markets Act.
Pires Alves’s recommendation fed the tribunal-level escalation now confirmed by Thomson de Andrade. The April 23 vote converts what had been a procedural revival into a formal administrative process — the step that allows CADE to gather evidence, issue subpoenas, and ultimately levy fines or remedies.
What Google said, and what Brazilian publishers want
Google’s statement after the vote characterised CADE‘s decision as reflecting “a misunderstanding of how its products work.” The company said it would continue to engage with the authority. That is a more measured response than Google’s earlier stance on European regulators, and signals that the Brazilian case is being treated as manageable by Alphabet’s competition-policy team.
On the other side, the Associação Nacional de Jornais, Brazil’s national newspaper association, has been pushing publicly for regulatory action that would oblige Google to negotiate compensation agreements with publishers. That model — the Australian News Media Bargaining Code — is the regulatory template the ANJ has cited most often.
The CADE decision does not itself impose such a regime. It creates the procedural conditions under which CADE could ultimately require it, either through a final ruling or through a negotiated commitment agreement.
The broader CADE posture on US tech
The Thursday vote sits inside a broader regulatory pattern. CADE has separately been investigating, since August 2024, whether Amazon, Google, and Microsoft should have formally notified antitrust authorities of their partnerships with AI startups.
CADE has also, in its published guidance, floated the idea of voluntary regulatory sandboxes for pricing algorithms and other AI systems — a framework that would let companies test AI deployments without incurring full antitrust liability.
The net posture is a Brazilian antitrust agency that is simultaneously inviting AI deployment and investigating how AI interacts with dominant market positions. Thursday’s Google decision is the first time those two threads have been formally joined in a named case.
What to watch after the Brazil CADE Google AI escalation
Three things now matter. The first is the pace of evidence-gathering. A formal administrative process at CADE typically runs two to four years.
The earliest plausible final decision is therefore late 2027 or 2028 — after Brazil’s October 2026 presidential election and a potential change of government.
The second is whether Google proposes a negotiated commitment agreement before that point. The company has used that mechanism in other jurisdictions to close investigations in exchange for behavioural remedies short of fines. A Brazilian commitment agreement that covered AI search features could become a reference template regionally.
The third is the read-across to other Latin American regulators. Mexico’s Cofece and Colombia’s SIC have both been tracking the European DMA investigation. A CADE case that explicitly places AI in scope, and that produces public evidence over the coming year, will give them a ready-made template if they choose to open their own proceedings.
For Alphabet and for investors reading Brazilian tech regulation, the signal is specific. The country’s competition authority has chosen not to accept the 2024 closure of this case, and has chosen to include AI in what it now investigates.
That is not a ruling. It is a decision that makes a ruling possible — and it lands just as the EU, the UK, and the US Department of Justice are all actively reshaping the legal perimeter around AI search.
Related coverage: Brazil 2026 mining and regulatory guide • EU reviews four Brazil critical mineral projects • USA Rare Earth acquires Serra Verde in Brazil