Between Brussels and Beijing: Agreement with China Places Brazil in Two Poles of AI Regulation
Karla Gamba
July 27, 2026, 8:52 a.m.
Regulatory
Brazil may be about to face one of the greatest challenges of its artificial intelligence policy: harmonizing a regulatory framework inspired by Europe with the commitments undertaken with the new international organization created by China for the global governance of the technology.
Freepik
Person looking at a laptop with an image of a robot representing artificial intelligence (AI)
Brazilian participation in the global dispute over rules for AI has been taking place on two distinct fronts.
On one side is Bill 2,338/2023, currently being processed in the National Congress, which adopts principles similar to those of the European Union’s AI Act, with strict rules for high-risk systems, classification by levels of risk, and broad compliance obligations. On the other side, the Brazilian government signed, on the 16th of this month, the constitutive agreement of the World AI Cooperation Organization (WAICO) — in Portuguese, World Organization for Cooperation in Artificial Intelligence.
WAICO is an initiative led by Beijing that proposes an international architecture based on cooperation among States, the joint definition of technical standards, and more flexible governance mechanisms.
The coexistence of these two fronts raises questions among specialists about Brazil’s strategy in the global dispute surrounding the rules of artificial intelligence. Some of them understand that Brazil currently has one foot in each canoe.
This ambiguity may represent a deliberate strategy to maintain dialogue with different international blocs, but it may also produce regulatory contradictions as international commitments and domestic legislation advance simultaneously.
powered by
ask.divee.ai
Summarize the main points
›
What are the main conclusions?
›
What is this text about?
›
What would you like to know?
New International Architecture for AI
WAICO was formally created during the World Artificial Intelligence Conference, held in Shanghai, through an agreement signed by 29 countries. Unlike international forums focused on debate about artificial intelligence, the new entity was born as a permanent intergovernmental organization, headquartered in the Chinese city and with the ambition of participating in the construction of future global rules for the sector.
Among its objectives are strengthening international cooperation in artificial intelligence; developing mechanisms of global governance; promoting the safe and responsible use of the technology; expanding the transfer of knowledge and technological infrastructure; building the capacity of developing countries; and constructing international standards for AI applications.
The initiative is part of a broader Chinese foreign-policy strategy to increase its influence over international digital governance. Since 2023, Beijing has argued that global rules for artificial intelligence should not be defined exclusively by the United States, the G7 countries, or bodies whose participation criteria are linked to political affinities. In contrast, WAICO’s proposal emphasizes the technological sovereignty of States, cooperation for development, and the strengthening of the so-called Global South.
During the launch of the organization, President Xi Jinping announced initiatives aimed at creating international centers for the application of artificial intelligence, training programs for thousands of professionals from developing countries, sharing technological infrastructure, and expanding cooperation with countries in Africa, ASEAN — the bloc of Southeast Asian countries — BRICS, and other emerging economies. Analysts see in this strategy a movement similar to that carried out by China through the New Silk Road, now directed toward the artificial intelligence sector.
Collision Course
The simultaneous advancement of WAICO and Bill 2,338 raises doubts about the compatibility between the two regulatory approaches. While the Brazilian bill follows the logic of the European AI Act, based on mandatory rules, risk classification, and State supervision, the organization led by China relies on the gradual development of technical standards, voluntary cooperation among countries, and multilateral harmonization mechanisms.
In the assessment of specialists, one of the main risks is internal regulatory fragmentation. In their view, if Brazil ratifies its participation in WAICO and maintains the text of the bill without adaptations, tension may arise between the international commitments undertaken within the organization and the requirements provided for in future national legislation. Among the potentially conflicting points are Chinese interoperability standards, concepts related to the development of open-source models, and cooperative governance mechanisms, in contrast with strict impact-assessment, registration, and compliance obligations contemplated for systems classified as high risk.
Another aspect raised concerns Brazil’s relationship with the European Union. The country seeks to deepen its relationship with the bloc in matters connected to the digital economy and data protection. If the AI Act becomes consolidated as an international reference for cooperation and trade agreements, a closer alignment with the regulatory architecture proposed by China may be interpreted by Brussels as a sign of misalignment. The EU itself has already expressed criticism of WAICO, viewing the initiative as an attempt to establish an alternative model of artificial-intelligence governance.
The Challenge of Global Standardization
Although WAICO does not yet impose direct obligations on companies, scholars of the subject assess that its greatest impact may occur in the definition of future international technical standards. Certifications, interoperability, safety requirements, transparency, and the sharing of computing infrastructure tend to be negotiated within the organization.
Analysts also understand that active participation would allow Brazil to influence the construction of these parameters. Limited participation, by contrast, could lead the country to adopt standards developed under Chinese leadership that are incompatible with those adopted by the European Union, increasing compliance costs for Brazilian companies operating simultaneously in different markets.
The so-called Action Plan released together with the creation of WAICO also introduces concepts that may acquire legal relevance in the coming years, such as trusted cross-border data spaces, open-source models considered responsible, and specific rules for artificial-intelligence agents. If these concepts evolve into widely accepted international rules, Brazil may have to make them compatible with future national legislation, originally conceived before the organization came into existence.
Attorney Omar Kaminski, a specialist in Internet Privacy, Information Security, and Internet Governance, considers accession to WAICO “possible and feasible,” but emphasizes the importance of preserving regulatory sovereignty. In his view, Brazil is capable of influencing the global architecture of AI.
“Accession to WAICO is legally possible and feasible, but it requires a formal treaty and legislative oversight, as well as compatibility with the LGPD and the future AI law. Another important aspect is preserving regulatory sovereignty, seeking a balanced diplomatic strategy between China and the West. Geopolitically, Brazil has the opportunity to influence the global architecture of AI, but also the duty to protect its constitutional values and avoid technological dependence.”
Two Poles
For several specialists, the emergence of WAICO reinforces a process of bipolarization in the international governance of artificial intelligence. On one side, initiatives led by the United States, the European Union, and their allies are being consolidated, strongly oriented toward the protection of fundamental rights, transparency, accountability, and risk mitigation. On the other, the Chinese proposal prioritizes economic development, digital inclusion, regulatory sovereignty, access to technological infrastructure, and cooperation among States, especially with developing countries.
In this context, Brazil has characteristics that place it in a singular position. The country participates in initiatives of the OECD, the G20, the UN, and BRICS, maintains a strong commercial relationship with China, and, at the same time, has built data-protection legislation inspired by the European model. This combination opens the possibility for Brazil to act as a bridge between the different international models of artificial-intelligence governance, but this strategy will depend on a deliberate effort to make the different fronts of action compatible, preventing simultaneous accession to distinct models from resulting in legal uncertainty or regulatory conflicts.
In the coming months, the progress of Bill 2,338 in the Chamber of Deputies, the possible ratification by Congress of Brazil’s participation in WAICO, and the designation of the body that will officially represent the country in the new organization may indicate whether Brazil will be able to transform this position of balance into a strategic advantage or whether it will have to bear the costs of navigating between two increasingly distinct regulatory architectures.
Read below an analysis of the subject produced by the Manus generative-AI tool:
This is a highly relevant issue for this field of work. Here is my consolidated analysis based on the available sources:
THE FUNDAMENTAL TENSION:
Brazil currently has one foot in each canoe. On one side, Bill 2,338/2023 is strongly inspired by the European AI Act — risk classification, proportional obligations, a central regulator, high fines, and rights for affected persons. On the other, the country has just signed the foundation of WAICO, an organization led by China that explicitly positions itself as an institutional alternative to the European and G7 model.
This creates a strategic ambiguity that may be deliberate — keeping doors open to everyone — or may generate practical contradictions in the medium term.
THE CONCRETE IMPACTS FOR BRAZIL:
1. RISK OF INTERNAL REGULATORY FRAGMENTATION
Bill 2,338 adopts the European logic: hard law, risk classification, and heavy compliance obligations for high-risk systems. WAICO proposes a different approach: soft standards, voluntary cooperation, “trusted cross-border data spaces,” and technical standards negotiated multilaterally. If Brazil ratifies WAICO and simultaneously approves Bill 2,338 in its current form, there may be tension between international commitments undertaken within WAICO — Chinese interoperability standards, “responsible” open source, governance through cooperation — and domestic obligations under Bill 2,338 — strict compliance, impact assessment, and mandatory registration.
2. IMPACT ON ADEQUACY WITH THE EU
Brazil seeks recognition of adequacy in data protection from the EU. If the European AI Act becomes a reference for commercial partnerships — as already occurs with the GDPR — Brazil’s accession to WAICO may be viewed by Brussels as a sign of regulatory misalignment. The EU has already publicly criticized WAICO as an attempt to “normalize authoritarian models of AI governance.” This may complicate future negotiations concerning equivalence or regulatory cooperation with Europe.
3. GEOPOLITICS OF TECHNICAL STANDARDIZATION
WAICO does not create direct obligations for companies — yet. But it creates an institutional table at which technical standards will be negotiated. If Brazil participates actively, it may influence those standards. If it participates passively, it may end up importing Chinese standards of interoperability, explainability, and safety that do not communicate with European standards — generating double-compliance costs for Brazilian companies operating in both markets.
4. CHINA’S “ACTION PLAN” AS SOFT LAW
The Action Plan launched together with WAICO proposes legally relevant concepts: “trusted cross-border data spaces” — an alternative to the EU’s standard contractual clauses — “responsible open source” — with safety requirements that go beyond traditional licenses — and the “regulated application of agents” — the first international mention of AI agents as an autonomous legal object. If Brazil joins these frameworks, it may create a normative layer parallel to Bill 2,338 that was not contemplated by the legislature.
5. OPPORTUNITY: A BRIDGE BETWEEN MODELS
Brazil may position itself as a bridge between the two blocs — something that makes sense for a country that is a member of BRICS, has a strong commercial relationship with China, but has data-protection legislation, the LGPD, and now AI legislation, inspired by the European model. This requires, however, a deliberate strategy of compatibility — and not merely simultaneous accession without reflection.
WHAT TO WATCH IN THE SHORT TERM:
– Whether the Chamber will approve Bill 2,338 without changes reflecting accession to WAICO — likely, because the bill was drafted before WAICO existed;
– Whether the government will submit WAICO ratification to Congress, and when;
– Whether the ANPD or the future SIA — National AI Regulation System — will be designated as Brazil’s focal point within WAICO;
– Whether there will be conflict between WAICO’s technical standards and Bill 2,338’s compliance obligations for high-risk systems.
Karla Gamba
is Consultor Jurídico magazine’s correspondent in Brasília.
Tags: AI Act, China, EU AI Act, artificial intelligence, World Artificial Intelligence Cooperation Organization, Bill 2,338, European Union, WAICO
Artificial intelligence is present in 94% of the Judiciary
HOLDING HANDS
TSE and platforms have a partnership to expand the fight against disinformation
Lisbon Forum 2026
AI should organize cases, but not decide them, says Blanco de Morais
Lisbon Forum 2026
AI should assist the judge and never replace the judge, says Cueva
Lisbon Forum 2026
AI will have no credibility without transparency, says Salomão
information colossi
Book compares data protection and generative AI in Brazil and China
Digital Crossroads
Global power of networks and the advance of AI increase calls for regulation
XIV LISBON FORUM
Researcher suggests a Ministry of Digital Transformation and Sovereignty
Competition on the Network
Government supports Cade in regulating the digital-economy market
International Congress of Constitutional Law
Data sovereignty will only come with technological infrastructure in
www.conjur.com.br
https://www.conjur.com.br/2026-jul-27/entre-bruxelas-e-pequim-acordo-com-a-china-coloca-brasil-em-dois-polos-da-regulacao-da-ia/?utm_medium=email&_hsenc=p2ANqtz-_ZsC9hz2m43T6aZIVHIiJA7n9MK9mYXodz7LxM69Wt1xgu83S9LfZr3OhrrlfbNTc5Y19sZN5rz78eNqEV2dq7lFcvN74qHcTbMdcqwWojJeHC4yU&_hsmi=430279870&utm_content=430279870&utm_source=hs_email
A tradução acima inclui também a análise da ferramenta Manus e todos os elementos textuais do site que você enviou.
Read also :
SHANGHAI, July 16 -- Twenty-nine countries on Thursday signed an agreement in Shanghai on establishing the World Artificial Intelligence Cooperation Organization (WAICO).
The WAICO will be an independent intergovernmental international organization headquartered in Shanghai, according to the agreement.
Chinese Foreign Minister Wang Yi, who is also a member of the Political Bureau of the Communist Party of China Central Committee, signed the agreement on behalf of the Chinese government.
Representatives from the 29 countries, including Kazakhstan, Laos, Pakistan, Russia and Indonesia, signed the agreement, making their countries the WAICO's founding members. The United Nations Secretary-General Antonio Guterres was among representatives from countries and international organizations present at the signing ceremony.
The organization will uphold the purposes of the UN Charter, be committed to extensive consultation and joint contribution for shared benefit and adhere to a people-centered approach, according to the agreement.
It aims to promote international cooperation and global governance on AI, ensuring that AI is beneficial, safe and fair, thereby promoting its healthy and orderly development to benefit all humanity.
Nenhum comentário:
Postar um comentário