AI for wellbeing? What does it tell us about Latin America

Written by: María Carrasco Rey. She is a Chilean public policy specialist and director of Entramada, where she works on participation, social innovation, and community wellbeing. An Atlantic Fellow at LSE, she currently explores how public policy and technology can contribute to fairer and more sustainable societies.

Technology aims to improve people’s lives. AI also plays a role in this sense. Thus, to evaluate its ultimate purpose, the wellbeing concept may be useful. 

I use the well-being idea for AI initiatives. Taking the well-being definition from the capability approach (Alkire, 2005, 2015; Robeyns, 2017; Sen, 1984, 2004), I define a wellbeing AI initiative as one that aims to increase or protect people’s abilities by clearly improving, or lowering risks and hardships, in one or more parts of wellbeing. The initiative must also think about how these effects affect different people and groups, what they mean for future generations, and how they connect with other parts of life.

Following the capability approach, the focus is on what AI allows people to be and do. Each project can be judged by five things: the ability it affects; the wellbeing area (such as health, education, income, housing, environment, safety, social ties, civic involvement, or how time is used). 

Analyzing the database of public AI projects in Latin America and the Caribbean implemented by Latin American governments, I found that most AI initiatives are related to surveillance and are heavily concentrated in facial-recognition initiatives, which are classified as potentially negative for wellbeing. 

AI for Latam population surveillance: general findings 

1. The geographical distribution is highly concentrated. Brazil accounts for 315/1,048 (30.1%) of the database, Colombia for 265/1,048 (25.3%), and Mexico for 151/1,048 (14.4%). 
2. Facial recognition initiatives are not a marginal application: they account for 293/1,048 (28.0%) of all recorded initiatives. 
3. The three most common technologies for AI initiatives, facial recognition, chatbots, and anomaly detection, represent 621/1,048 = 59.3% of the database. 

These overall figures already indicate a tendency to develop AI technology related to surveillance.

 

Table 1: Descriptive Indicators of general findings

Initiatives analysed	1,048	Individual country-level records
Countries covered	26	Uneven coverage across countries
Brazil, Colombia and Mexico	731 — 69.8%	Nearly 7 out of 10 records come from three countries
Five leading countries	871 — 83.1%	Brazil, Colombia, Mexico, Chile and Argentina
Facial recognition	293 — 28.0%	Most frequently recorded AI technology
Chatbots	190 — 18.1%	Second most frequent technology
Anomaly detection	138 — 13.2%	Third most frequent technology
Three leading technologies	621 — 59.3%	High technological concentration
Executive branch	753 — 71.9%	Main institutional setting

 

AI for Wellbeing?

Key findings about well-being and AI:

1. 45% of the projects can clearly be linked to wellbeing, but 36% of that percentage is related to surveillance initiatives. Wellbeing is also linked to surveillance initiatives such as facial recognition, as their stated objective is to protect people’s lives and keep them safe from crime and violence. 
2. Only a small portion of AI initiatives relate to wellbeing. Following the analysis, 46% had no connection to wellbeing impacts, 11% had a small connection, 22% had a moderate connection, 19% were connected, and only 2% were strongly connected to well-being.
3. Some of the wellbeing-related initiatives might harm wellbeing. 26% of the surveillance AI initiatives could discriminate against vulnerable groups and communities already treated unfairly. 

 

Table 2: Well-being indicators

Concept	Relation figures	Percentage
Initiatives related to wellbeing	457/1,048	43.6%
Positive direction among wellbeing-related initiatives	287/457	62.8%
Negative direction among wellbeing-related initiatives	170/457	37.2%
Wellbeing + facial recognition + control	169/1,048	16.1% of all initiatives
Control among facial recognition initiatives related to wellbeing	169/177	95.5%
Negative direction within the triple intersection	153/169	90.5%
Triple intersection among all negative wellbeing-related initiatives	153/170	90.0%

Source: Author’s elaboration.

 

It seems that when governments use AI, they tend to focus more on public security systems to protect people from violence but, in doing so, may create serious problems for certain groups. The main question is whether this type of technology, which costs governments millions, achieves its goal of protecting people by reducing violence and insecurity. 

LATAM wellbeing excuse: AI for surveillance and facial recognition

Key country-level findings:

1. Brazil is the clear outlier. 70% of initiatives combine wellbeing, facial recognition, and surveillance. This intersection represents 37.5% of Brazil’s initiatives. 
2. Colombia has the lowest wellbeing share among countries with larger samples, at 26.4%. 
3. Mexico and Chile both stand at around 9% in the triple intersection of wellbeing, facial recognition, and surveillance. 

 

Table 3: Countries’ indicators

 

Country	Total AI initiatives	Well-being-related initiatives, n (% of country total)	Facial recognition, well-being and control intersection, n (% of country total)
Brazil	315	179 — 56.8%	118 — 37.5%
Colombia	265	70 — 26.4%	9 — 3.4%
Mexico	151	57 — 37.7%	14 — 9.3%
Chile	87	50 — 57.5%	8 — 9.2%
Argentina	53	26 — 49.1%	3 — 5.7%
Peru	46	23 — 50.0%	4 — 8.7%
Uruguay	35	13 — 37.1%	1 — 2.9%

Source: Author’s elaboration.

The hope landscape: What should governments ask before deploying AI?

Last week, UN human rights chief Volker Türk warned that AI is a greater existential risk to every aspect of our lives. And yes, it is true. The autonomous AI drones that killed three people in Ukraine should warn us about the power of this technology. 

We live in a time when democracy is rare. According to the latest V-Dem Institute report (2026), only 26% of people live in democracies. We all need to protect democracies because it is the only system that we know so far in modern societies, where people can freely choose their own paths in a peaceful manner, which is essential for a good life. But if people keep feeling that governments focus more on controlling them than on helping them live well and with dignity, how can we expect to keep supporting democracies?

A new way of thinking about government AI initiatives is to identify the capabilities expected to expand through the initiative. The health sector, for example, has many positive examples of using AI to expand people’s possibilities to do and be, either by preventing disease, or providing more efficient services. In this sense, adopting an early-intervention perspective, such as the one implemented by the Government of Victoria, could be a good example of how and when to implement AI initiatives: before the public problem is unleashed, by preventing it. That is how we ensure people’s ability and opportunity to live well.

Methodology box

I turned the capability approach into four parts: whether a project involved watching or control; whether it was linked to wellbeing; how strong that link was on a five-point scale; and the possible effect direction from -2 to +2.

I tested the coding rules using cases from Chile. ChatGPT then used the same rules on the other projects based on their names, organizations, sectors, keywords, and descriptions. I later checked and confirmed the AI-assisted coding by hand, fixing any differences, especially in surveillance and security cases.

After leaving out two OAS records, the analysis includes 1,048 projects from 26 countries. The scores show possible links and effects based on available information; they do not prove real impact. The database is incomplete and not equally balanced across countries.