The introduction of Digital Tribulations, a series of intellectual interviews on the developments of digital sovereignty in Latin America, can be read here. 
For the last interview in Mexico, we dive into the more institutional perspective with the work, value and activities of the newly created Agency for Public Innovation of Mexico City. After attending a public consultation on the 20-years digital sovereignty plan, I managed to talk online to the head Angel Tamariz on February 20, 2026. The Agency is currently led by Emiliano Calderón.
***
What does the Agencia Digital de Innovación Pública (ADIP) do?
The Digital Agency for Public Innovation is an entity of the Mexico City government that was created under the administration of Claudia Sheinbaum, who is now president. The agency’s aim is to bring together, in a single entity, not only the use of technology in public service and in the relationship between government and citizens, but also the authority to facilitate that communication and interaction between citizens and government.
This goes beyond simply digitizing services; it has to do with simplifying procedures and citizen services. That is why we have the authority to evaluate any proposal for regulatory improvement. Normally, we propose something, and it is always implemented, but if we need the authority to require procedures to be simplified, we have it. And then, where appropriate, comes digitization. Our goal is to be a technological arm of the government, but we are not an internal technology department. Our main priority is a policy for leveraging technology within the relationship between citizens and the government, which carries implications across the board.
What is the concept of digital sovereignty that you use?
For ADIP, digital sovereignty means being able to make decisions independently, without influence or interests external to the government. It means having the freedom to make digital changes, strategies, and implementations while thinking only of the general interest of citizens. We have applied this in many ways. A very clear example is software development: practically all the systems we build at the agency are developed by our own team of developers, with the goal of knowing the code, understanding the structure of what is being developed, and having governance over the data that is generated, which is also one of the key priorities.
Now, this does not mean that if we hire a company for a development project, we are violating this principle, as long as the rules are very clear regarding delivery of the code, ownership of the code, and governance of the data. Right now, several ministries are suffering from a kind of technological lock-in, a software lock-in, where in order to make modifications or updates, payment is required; if payment is no longer made, the source code is unavailable, making updates complicated. There is a need to strengthen this area, and we are doing so. Over the past seven years, it has been increasingly reinforced.
But this policy of digital sovereignty in software development entails major responsibilities. We, as public officials, are temporary; we are passing through, ephemeral. Software does not necessarily have to be. Therefore, the interest in technological sovereignty in software development carries with it the responsibility to ensure that the survival of the software does not depend on individuals, but on the documentation that has been left behind. As for the sovereignty of the infrastrcuture, ideally we would build our own servers, but that capacity does not yet exist in Mexico. However, we do have our own data center. The administration of the servers where the information is stored is in Mexico and belongs to the Government of Mexico City. The agency has two very important data centers where we manage everything.
So, what we want to maintain in terms of sovereignty is the ability to make decisions about the software, own the source codes, have the ability to update the systems, have absolute governance over the data we manage, and ensure that the systems live on servers administered by the city government itself.
In this discourse, is there a distinction between Mexican companies and foreign companies? Is there also an open-source dimension?
Yes, we always review the use of free software, or at least open-source code, in everything we use. Sometimes it is not possible. As much as we have fought to have alternatives to instant messaging, we know that social networks exist, as do commercial messaging systems that everyone uses.
And here comes a very interesting discussion: our principles are in favor of free culture and free software, but what is also fundamental is what is practical for the population. If we wanted to use another messaging system but only a small percentage of the population had it, we would not be fulfilling our main function, which is to govern for everyone. So we use commercial messaging systems that, unfortunately, we have not been able to replace with open-source alternatives. And the other level is that when we hire companies, we ideally seek companies that are based in Mexico City, but if not, at least in Mexico, because this also has to do with how far the data goes and who has it.
The other day I participated in the public consultation for the digital sovereignty plan, and I thought the idea of a long-term 20-year plan was very important — something I did not find in other countries. Could you explain some concrete projects and implementation challenges?
A very clear example is that, amid this entire wave of artificial intelligence, we are developing citizen-assistance chatbots and chatbots for tourists and cultural assistance. They are very nice projects because people receive information very quickly. But we have faced very significant challenges. The tourism-cultural chatbot, which we also want to launch for the World Cup, is intended to be conversational, using generative artificial intelligence. That requires very powerful natural language engines and very powerful servers.
We have explored the possibility of installing language models on our own servers, but there we face an infrastructure challenge: if we install them on our servers, they are less powerful and have fewer capabilities. So in recent months we have been balancing the need for something functional without necessarily contracting all services from commercial companies. How do we strike a balance between strengthening sovereignty and investing in servers that are very expensive — and that are currently rising in price because of scarcity — and determining to what extent we pay for services from these language models in order to balance functionality with sovereignty? We have already made a decision: a balance between paying for some APIs and installing certain parts on our own servers, with the perspective of depending less and less on those services over time.
Another challenge is strengthening the agency’s own knowledge in these areas, because the research and innovation department is new. One of its main tasks is the use of artificial intelligence in public service. As we generate that knowledge and explore at the pace at which the field is advancing, we are caught in that dilemma between functionality and the constant challenge of pursuing technological sovereignty.
Another issue that interests me a lot is digital public infrastructure. The case of Brazil with PIX seems revolutionary to me. Is the agency or the Mexican government doing anything in that regard, including on digital identity?
Yes, PIX is very important. Here in Mexico, similar efforts have been made by the federal government together with the Bank of Mexico, which is autonomous. DIMO exists, and so does CoDi, but in reality they have not achieved the level of adoption among the population that we would like, in order to stop depending on cash and be able to carry out more transactions digitally. What has worked best is SPEI, which is another transfer protocol where you enter the other person’s bank account, but I feel it is still not as fluid or agile as PIX. That is why DIMO was created, which works with your phone, but like CoDi, it has not achieved mass use.
As for digital identity, at the federal level, the most official form of identity today is the INE voter card. However, it has several disadvantages; the most obvious is that in order to vote you have to be 18 years old, so all children lack that identity. The other thing that exists — and this is for all citizens regardless of age — is the Unique Population Registry Code (Clave Única de Registro de Población, CURP) which is an 18-character number. The federal government has a project to turn the CURP into an identification document and also link it with biometric data to guarantee that it is unique for each person. This is intended to facilitate government procedures and avoid relying solely on the INE credential.
In Mexico City, we have made some progress in terms of digital authentication for procedures: Llave CDMX. It is a way to authenticate yourself, but it is not a digital identity, because it cannot guarantee that the person behind the user account is who they say they are. What we are implementing this year is Llave Plus or Llave Verificada, which is strengthened and linked to biometrics on an optional basis. It is not that everyone has to have it, because there is a universe of procedures where it does not matter who is behind the screen. If you are going to pay the water bill, whether you are Marco or Ángel does not matter. But there are procedures where I do have to make sure that the person claiming to be someone is really that person — for example, applying for a driver’s license. So we carry out remote biometric verification: from home, you can complete a verification process to confirm that it really was you who created that identity.
And there would be a third level, not only when you create the Llave account, but when you are going to carry out the procedure, where there would again be biometric validation. With Llave Verificada, we believe this is the first concrete step toward digital identity at the Mexico City level, while at the federal level it will be the biometric CURP.
Is there dialogue between the agency and other Mexican institutions, and between the Mexican government and other Latin American countries?
With some organizations, yes, but perhaps less than desired. We have engaged especially in discussions around the digital divide. But on issues of sovereignty, more dialogue is needed; and with other Latin American countries, definitely much more is needed. There have been some exchanges to discuss issues such as artificial intelligence or accessibility of digital portals for people with disabilities, but these are more practical matters, not as deep in terms of the philosophical and political aspect of what technological sovereignty implies.
I think those dialogues would be worthwhile. One of the problems is that I believe we are in the process of internally strengthening the sovereign aspect of implementations within Latin American governments. But I think this will happen quickly, so that once it is more or less resolved — in terms of data governance, software development, and our own servers — we can turn toward other Latin American countries and think about how, as a region, we can strengthen the region’s sovereign dimension, especially in areas such as infrastructure or language models for artificial intelligence.
Do you think there is more discourse around digital sovereignty in this government compared with previous governments?
I think that starting with the previous government, greater relevance was given to the issue of digital sovereignty, without a doubt, here in Mexico City. Before, it was not an issue that mattered. All primary and secondary schools in the city, if they acquire a computer lab, what operating system do those computers have? Well, they have a proprietary operating system installed, which will later depend on paying licenses in order to use it — and nobody questions that.
The first challenge is for there to be computers, internet, and electrical infrastructure; and only then do we begin discussing free software and sovereignty. However, we are launching the Mixtli Digital project in Mexico City, which means that over six years, starting last year, all public primary and secondary schools will have a digital classroom with computers, tablets, digital screens, Arduinos, and robotics kits. We, as the agency, were technically responsible for determining what would be acquired and under what conditions. And although the computers do not run only Linux, all public primary and secondary schools in the city will have both Windows and Linux.
That is already a minimal platform so that different organizations and associations can promote the use of Linux in schools and reduce that dependency. So I believe this government represents a deepening of the concern for digital sovereignty that began in the previous six-year administration.
What would be a pragmatic outlook for the coming years? What are the areas where more work can be done?
I think one of the most relevant issues for this government, and one that we have begun to take responsibility for, is combating the digital divide. All reflection on sovereignty must go hand in hand with guaranteeing access for everyone, because that remains a problem in Mexico City. One might think that here everyone already has a cell phone, everyone has internet, everyone knows how to use it — and that is not the case. It is definitely not the case. Perhaps we have the lowest levels of digital divide in the country, perhaps; but today I can affirm that any digital solution made available to people does not guarantee that everyone will be able to use it. It is not a democratic solution, and therefore there must be a proactive stance from the government to guarantee those rights.
The head of government told us that, as a digital agency, we should not remain “in the cloud”; we have to go out into the territory to see what people’s reality is, so that the use of technology does not reach only some people and, even worse, does not generate greater inequalities or new inequalities. Our perspective for the future is that the agency should take a leading role in analysis, proposals, and actions to combat the digital divide, making it possible to guarantee digital rights and access to government services for the entire population.