Greater Autonomy or Deeper Dependence?
.png)
The
Alliance was created against a backdrop of highly concentrated technological
capabilities. The United Nations Conference on Trade and Development (UNCTAD)
highlighted this problem in its Technology and Innovation Report 2025,
warning that many developing economies face difficulties accessing the
infrastructure and knowledge needed to participate in AI development (UNCTAD,
2025). In Latin America and the Caribbean, these challenges are compounded by
substantial differences in technological capabilities across countries (ECLAC,
2025).
The new
Alliance seeks to address these challenges through cooperation, building on
existing resources and helping countries develop shared capabilities.
Argentina
and Paraguay declined to endorse the declarations establishing the IAIA,
raising questions about the priorities that shape international partnerships on
AI.
Data, Models, and Computing: What Does the IAIA Propose?
The
creation of the IAIA was first agreed upon on November 3, during the First
Ibero-American Ministerial Meeting on Digital Transformation and Artificial
Intelligence, and formally adopted by the Summit two days later. The Alliance
is designed as a cooperation mechanism with three main components (First
Ibero-American Ministerial Meeting, 2026):
- The gradual establishment of an
Ibero-American Data and Interoperability Space, using open standards and
federated architectures to enable secure information exchange while
respecting each country's ownership and control of its data and
infrastructure under national law. The political significance of this
approach connects with issues I have previously explored in Data
Geopolitics: Power, Infrastructure, and Dependence, particularly the
relationship between data governance, interoperability, and state capabilities.
- The collaborative development
of AI models and systems adapted to Spanish, Portuguese, and, over time,
Indigenous languages and other official languages spoken across
Ibero-America. This effort could build on existing initiatives such as
ALIA, AMALIA, LATAM-GPT, and SoberanIA, developed in Spain, Portugal,
Chile, and Brazil, respectively (SEGIB, 2026).
- Access to high-performance
computing resources by connecting existing infrastructure and making it
available for shared projects of public interest, without requiring the
construction of new regional facilities. This is particularly relevant
given the considerable disparities in computing capacity across the
region.
The
Alliance seeks to bring existing capabilities together and make them more
widely available, reducing duplication and improving access for countries with
fewer technological resources.
Technological Sovereignty: Access Versus Control
The Madrid
Ministerial Declaration defines digital sovereignty in terms of countries'
ability to make decisions about, develop, adapt, govern, and use technologies,
data, and critical infrastructure in accordance with their public interests and
legal systems. It also recognizes the role of cooperation and public- and
private-sector participation, according to national priorities and capabilities
(First Ibero-American Ministerial Meeting, 2026).
I find this
definition useful because it allows us to assess technological sovereignty in terms of concrete capabilities. It also makes an important distinction between
having access to a technology, being able to use it, and controlling the
decisions that shape how it operates.
As I
discussed in The Geopolitics of Cloud Computing: Infrastructure,
Sovereignty, and AI Power, access to technological services can expand what
states are able to do, but it does not guarantee control over the infrastructure on which those services depend. The same applies to AI models.
The ability to adapt systems developed by others may give countries more
technological options, but their autonomy will also depend on licensing terms,
access to model components, update policies, and the availability of computing
infrastructure to run them.
With this
distinction in mind, I propose assessing technological autonomy by examining
which decisions countries can actually make and sustain, including when the
external conditions governing access to essential resources change.
The IAIA
could strengthen that autonomy if it helps countries develop their own
capabilities, broadens their technological options, and gives participants a
meaningful role in decisions about shared resources. But it could also improve
access to existing technologies, which would be a welcome development, while
leaving underlying relationships of dependence largely unchanged. That outcome
could deepen inequalities within the region.
The Madrid
documents establish a broad political direction, but they do not yet specify
the mechanisms that would allow us to determine which outcome is more likely.
Who Will Hold Power Within the Alliance?
In an
earlier analysis of AI and international development cooperation, I warned that
technological cooperation can reproduce inequalities when it fails to
strengthen countries' own capabilities. The new Alliance brings that concern
back into focus. The Ministerial Declaration acknowledges that Ibero-American
countries differ in their levels of digital development, computing capacity,
and institutional strength. It therefore calls for cooperation that is gradual,
voluntary, and flexible (First Ibero-American Ministerial Meeting, 2026).
Countries
joining the IAIA would take an active role and assume shared responsibilities,
contributing financial resources or other public assets according to their
circumstances. These arrangements will have to be developed within a common
governance framework (SEGIB, 2026).
This raises
a question that remains unanswered and that I consider critical: How will each
country's contribution of resources affect its influence over collective
decisions?
As I have
already argued, the IAIA could end up entrenching existing inequalities and
strengthening the leadership of a handful of countries if it fails to ensure
that members with fewer resources can participate effectively in
decision-making.
A similar
concern applies to technology companies. Their involvement may bring
infrastructure, investment, and specialized expertise, but that should not give
them greater authority to determine the IAIA's priorities, decide how rights
over jointly developed technologies are allocated, or shape the mechanisms
intended to prevent technological dependence.
Governance
rules will therefore matter far beyond the Alliance's internal administration.
They will influence how technological power is distributed among its
participants.
Argentina and Paraguay: Two “Defections” That Demand an Explanation
The
Ministerial Declaration includes explicit notes recording that Argentina and
Paraguay dissociated themselves from the document establishing the IAIA (First
Ibero-American Ministerial Meeting, 2026; 30th Ibero-American Summit, 2026).
Both countries also declined to endorse paragraph 10 of the Summit Declaration,
which formally establishes the Alliance. They expressed reservations about
other provisions as well, but those fall outside the scope of this analysis of
AI.
Unfortunately,
despite a thorough search, I have been unable to find official documents
explaining their objections to the IAIA. More troubling still, in Argentina's
case, a review of the government's decisions reveals a striking inconsistency.
On October
31, 2025, Resolution 282/2025 was published in Argentina's Official Gazette.
Through this resolution, the Secretariat of Innovation, Science and Technology
approved the Strategic Guidelines for Innovation, Science and Technology
2025–2027. The guidelines identify AI as a priority technology and promote
international cooperation, participation in international networks, and
scientific and technological partnerships as ways to strengthen national
capabilities (Secretariat of Innovation, Science and Technology, 2025).
On November
3, 2026, Argentina formally distanced itself from the Madrid Declaration
adopted at the First Ibero-American Ministerial Meeting on Digital
Transformation and Artificial Intelligence.
On November
4, 2026, the same Secretariat issued a statement about Argentina's
participation in the meeting. Its representative, Gerardo Machado,
Undersecretary of Information and Communication Technologies, had emphasized
the country's scientific and technological capabilities, its energy potential,
and the importance of working with the private sector. In his closing remarks,
he reaffirmed Argentina's willingness to participate in concrete international
cooperation initiatives involving research, talent development, infrastructure,
and AI applications (Secretariat of Innovation, Science and Technology, 2026).
Yet on
November 5, just one day later, Argentina once again distanced itself from the
IAIA, declining to endorse the paragraph of the Summit Declaration establishing
the Alliance.
I believe
this sequence exposes an inconsistency in the Argentine government's approach
to technological cooperation that deserves a public explanation. Both
Argentina's Strategic Guidelines for Innovation, Science and Technology
2025–2027 and the areas of cooperation identified by Machado closely match the
objectives of the new Alliance.
I recognize
that a government can support international cooperation while rejecting a
particular arrangement if it believes that arrangement conflicts with its
interests or priorities. But the Argentine authorities have offered no
explanation of what those reasons might have been, assuming there were any.
Nor is it
enough, in my view, to attribute the decision to the government's general
preference for deregulation. Participation in the Alliance is voluntary, the
initiative respects national legal frameworks and priorities, and it allows
private-sector involvement. Furthermore, the Ibero-American Data Protection
Standards, developed by the Ibero-American Data Protection Network (RIPD), and
the Ibero-American Standards for Identification and Interoperability in Digital
Government Services, developed through the work of the Latin American Center
for Development Administration (CLAD), were welcomed by the Ministerial
Declaration. The document expressly uses the term "welcome," and both
sets of standards are non-binding (First Ibero-American Ministerial Meeting,
2026).
This leaves
an unanswered question: why would Argentina distance itself from an initiative
whose stated objectives are consistent with its own technological policy
priorities?
In
Paraguay's case, I have also been unable to find a specific official
explanation. Its recent international cooperation initiatives, however, provide
some context. In May 2026, Paraguay signed a cooperation agreement with Taiwan
to develop a sovereign AI computing center (MITIC, 2026a).
Pax Silica: A Geopolitical Clue, but Not a Sufficient Explanation
Argentina
and Paraguay both participate in Pax Silica, an initiative launched by the
United States in December 2025 to strengthen strategic supply chains and
infrastructure linked to AI. Its agenda covers critical minerals, energy,
semiconductors, advanced manufacturing, and economic security, bringing
governments and private-sector actors together (U.S. Department of State,
2026).
Argentina
signed the main Pax Silica Declaration at the initiative's second summit, held
in Washington on June 25–26, 2026, alongside Chile, Costa Rica, El Salvador,
and Panama.
Paraguay
signed on October 23, just thirteen days before the Madrid Summit. Its Ministry
of Foreign Affairs presented the agreement as an opportunity to use the
country's hydroelectric potential and critical mineral resources to develop
advanced AI infrastructure in the Americas (Ministry of Foreign Affairs of
Paraguay, 2026).
However,
six Latin American countries had signed the main Pax Silica Declaration by the
time of the Madrid Summit. Participation in that initiative therefore cannot,
by itself, explain why Argentina and Paraguay objected to the IAIA.
I do not
rule out a geopolitical dimension to their decisions. But I believe that any
such explanation requires a more careful analysis, beginning with greater
transparency about why these two countries chose to distance themselves from
the Ibero-American agreement.
For now,
the available evidence shows that they made different choices regarding
cooperation arrangements that are not inherently incompatible. What motivated
those choices remains unclear.
The Real Test: Turning Cooperation into Power
The
creation of the Ibero-American Artificial Intelligence Alliance gives countries
in the region an opportunity to expand their digital capabilities through
cooperation. Its initial design also acknowledges the inequalities that already
exist among them.
Even so,
the political significance of that cooperation will depend on the governance
rules adopted, the distribution of shared resources, and whether countries can
participate meaningfully in collective decisions.
Argentina's
and Paraguay's decisions to distance themselves from the Alliance suggest that
international technology partnerships are shaped by more than technical needs. Political
and institutional priorities appear to be involved in both cases, although I
have found no documents confirming their specific motivations.
We will
have to see how the IAIA develops before we can judge whether its promises of
cooperation translate into new capabilities for participating countries, and
whether those capabilities give them more options when dealing with the actors
that currently control strategic technologies.
In a world
marked by unequal technological dependencies, countries' autonomy will depend
not only on the resources they can access and use, but also on their ability to
determine the terms on which they participate in the systems that provide those
resources.
Ultimately, the IAIA's value will be measured by its ability to bring about real change within the region and by the influence Ibero-America manages to gain in global decision-making.
References
Economic
Commission for Latin America and the Caribbean (ECLAC). (2025). Índice
Latinoamericano de Inteligencia Artificial (ILIA) 2025. United Nations. https://www.cepal.org/es/publicaciones/82514-indice-latinoamericano-inteligencia-artificial-ilia-2025
United
Nations Conference on Trade and Development (UNCTAD). (2025). Technology and
Innovation Report 2025: Inclusive artificial intelligence for development.
United Nations. https://unctad.org/publication/technology-and-innovation-report-2025
U.S.
Department of State. (2026, June 26). Pax Silica [Fact sheet]. https://ebs.publicnow.com/view/4BFCA069838BA31827FC1B5D6217C610EA5B54C7
First
Ibero-American Ministerial Meeting on Digital Transformation and Artificial
Intelligence. (2026, November 3). Declaración de Madrid.
Ibero-American General Secretariat. https://segib.org/wp-content/uploads/2026/11/Declaracion_Madrid_Transformacion_Digital_IA-ES.pdf
Ministry of
Foreign Affairs of Paraguay. (2026, October 23). Paraguay se suma a
la Declaración de Pax Silica. https://www.mre.gov.py/paraguay-se-suma-a-la-declaracion-de-la-pax-silica/
Ministry of
Information and Communication Technologies of Paraguay (MITIC). (2026a,
May 13). Alianza con Taiwán abre camino al desarrollo de la IA en Paraguay.
https://mitic.gov.py/alianza-con-taiwan-abre-camino-al-desarrollo-de-la-ia-en-paraguay/
Secretariat
of Innovation, Science and Technology of Argentina. (2025, October 30). Resolución
282/2025. Lineamientos Estratégicos en Innovación, Ciencia y Tecnología
2025–2027. Official
Gazette of the Argentine Republic. Published October 31, 2025. https://www.argentina.gob.ar/normativa/nacional/resoluci%C3%B3n-282-2025-419597/texto
Secretariat
of Innovation, Science and Technology of Argentina. (2026, November 4). Promovemos
la aplicación de la IA para impulsar la innovación y el desarrollo. https://www.argentina.gob.ar/noticias/promovemos-la-aplicacion-de-la-ia-para-impulsar-la-innovacion-y-el-desarrollo
Ibero-American General Secretariat (SEGIB). (2026, November
3). Los países iberoamericanos acuerdan en Madrid la Alianza Iberoamericana
de Inteligencia Artificial. https://segib.org/es/los-paises-iberoamericanos-acuerdan-en-madrid-la-alianza-iberoamericana-de-inteligencia-artificial/
30th
Ibero-American Summit of Heads of State and Government. (2026, November
5). Declaración de Madrid. Ibero-American General Secretariat. https://segib.org/wp-content/uploads/2026/11/Declaracion_de_la_Cumbre_de_Madrid-2026.pdf