OEA - CP- Resolución XIX VI 1 CIMT 22 GT1 GT2 8

INTER-AMERICAN COUNCIL FOR INTEGRAL DEVELOPMENT

MEETING OF WORKING GROUPS 1 AND 2 OEA/Ser.L/XIX.VI.1

OF THE TWENTY-SECOND INTER-AMERICAN CIMT-22/GT1-GT2/doc.8/26

CONFERENCE OF MINISTERS OF LABOR 7 May 2026

May 11 to 13, 2026 Original: Spanish Asunción, Paraguay OUT

COMES AND RECOMMENDATIONS

OF THE HEMISPHERIC WORKSHOP “ Artificial Intelligence, digital platforms and labor rights in the Americas”1

OUTCOMES AND RECOMMENDATIONS OF THE HEMISPHERIC WORKSHOP “Artiϐicial Intelligence, digital platforms and labor rights in the Americas”

INDEX

INTRODUCTION……………………………..………………………………………..………………………………………..…….. 2

1. MAIN IDEAS AND GENERAL CONSIDERATIONS 1.1. ArƟficial Intelligence, Employment and Skills ……………..………………………………..………..……..... 3 1.2. AI and emerging technologies for the benefit of the Ministries of Labor ……………………….... 9 1.3. Work on digital plaƞorms …………………………..…………………………………………….…………..………. 13

2. RECOMMENDATIONS 2.1. RecommendaƟons on AI, Employment and Skills ……………………………..……………………….…… 18 2.2. RecommendaƟons on AI for the benefit of the Ministries of Labor……………………………….… 20 2.3. RecommendaƟons on work on digital plaƞorms ………………..…………..………….………………..... 21 2.4. RecommendaƟons on cooperaƟon and for the RIAL/OAS ……………………………………..……….. 22

3. EVALUATION OF THE WORKSHOP....................................................................................... 23

ANNEXES

2.4. RecommendaƟons on cooperaƟon and for the RIAL/OAS ……………………………………..……….. 22

3. EVALUATION OF THE WORKSHOP....................................................................................... 23

ANNEXES

Annex No. 1 – Workshop’s Agenda ............................................................................................... 25 Annex No. 2 – List of parƟcipants ................................................................................................. 302

INTRODUCTION

The Hemispheric Workshop “ArƟficial Intelligence, digital plaƞorms and labor rights in the Americas” was held on September 11 and 12, 2025, in a hybrid form in Bogotá, Colombia, as part of the 2025-2027 Work Plan of the Inter-American Conference of Ministers of Labor (IACML) of the OrganizaƟon of American States (OAS). Accordingly, the event responded to the mandates established by the Ministers of Labor in the DeclaraƟon and Plan of AcƟon of Bogotá, adopted at the XXII IACML in 2024, addressing these priority topics.

The Workshop was an acƟvity of the Inter-American Network for Labor AdministraƟon (RIAL), coorganized by the Ministry of Labor of Colombia and the OAS, and involved delegaƟons from 29 Ministries of Labor from across the region, representaƟves of workers -grouped under the Trade Union Technical Advisory Council (COSATE)- and employers -grouped under the Business Technical Advisory CommiƩee on Labor MaƩers (CEATAL)-, the Ministry of Labor of Spain as an OAS observer country, and specialists from the Inter-American Development Bank (IDB), the Economic Commission for LaƟn America and the Caribbean (ECLAC), the OrganizaƟon for Economic CooperaƟon and Development (OECD), the InternaƟonal OrganizaƟon of Employers (IOE), the InternaƟonal Labor OrganizaƟon (ILO), the OAS Centers

Caribbean (ECLAC), the OrganizaƟon for Economic CooperaƟon and Development (OECD), the InternaƟonal OrganizaƟon of Employers (IOE), the InternaƟonal Labor OrganizaƟon (ILO), the OAS Centers of Excellence in Science and Technology -represented by CENIA (Chile’s NaƟonal Center for ArƟficial Intelligence) and AudacIA (Colombia)-, as well as the Fairwork Project of the University of Oxford, Johns Hopkins University, and the OAS.

In line with the mandates of the DeclaraƟon and Plan of AcƟon of Bogotá, the general objecƟve of the Workshop was to enhance the knowledge and strengthen the capaciƟes of Ministries of Labor and social actors to strategically address the challenges and opportuniƟes presented by technological advancements –parƟcularly AI– in the world of work, while promoƟng insƟtuƟonal innovaƟon, decent work, and regulatory frameworks aligned with new forms of employment. The specific objecƟves were:

• To idenƟfy, exchange, and analyze the impacts of AI on the world of work, parƟcularly regarding the transformaƟon of skills required for employment, and to explore strategies enabling Ministries of Labor to anƟcipate and respond to these emerging demands.

• To share and analyze experiences related to the use of AI and other emerging technologies across the various areas of intervenƟon within Ministries of Labor.

• To exchange and analyze policies and regulatory frameworks developed by Ministries of Labor in the region aimed at addressing work on digital plaƞorms from a rights-based perspecƟve.

• To develop policy recommendaƟons that guide Ministries of Labor in responding to the challenges and opportuniƟes associated with emerging technologies –especially AI– with a focus on their

the region aimed at addressing work on digital plaƞorms from a rights-based perspecƟve.

• To develop policy recommendaƟons that guide Ministries of Labor in responding to the challenges and opportuniƟes associated with emerging technologies –especially AI– with a focus on their impact on employment, their potenƟal to strengthen insƟtuƟonal capacity, as well as the growth of work on digital plaƞorms.

To achieve these objecƟves, the Workshop featured three themaƟc sessions: the first session on “AI, employment and skills: What can we expect and what should we do?”, a second session on “AI and3

emerging technologies for the benefit of Ministries of Labor”, and a final session on “Work on digital plaƞorms and labor rights”. Each session included guiding quesƟons to frame the discussions. Within the sessions, there were presentaƟons of demonstraƟve government experiences from ArgenƟna, Canada, Chile, Colombia, Jamaica, Mexico, Panama and Uruguay; as well as presentaƟons from AudacIA, IDB, CENIA, ECLAC, Fairwork, Johns Hopkins University, OECD, OAS, and ILO, and from workers and employers, represented by COSATE and CEATAL. There were also broad dialogue spaces where Ministries and representaƟves of workers and employers shared their reflecƟons and experiences.

On the second day of the event, a sub-group exercise was conducted, allowing parƟcipants to discuss in greater depth and collecƟvely answer the quesƟon of what lessons learned and policy recommendaƟons could be idenƟfied, based on the discussions held during the Workshop.

Four (4) sub-groups were organized (two on-line and two in-person), moderated by government delegates from ArgenƟna, Barbados, Ecuador, and Trinidad and Tobago. DelegaƟons from the 29

Four (4) sub-groups were organized (two on-line and two in-person), moderated by government delegates from ArgenƟna, Barbados, Ecuador, and Trinidad and Tobago. DelegaƟons from the 29 parƟcipaƟng Ministries, along with representaƟves of workers and employers, acƟvely parƟcipated in the subgroup discussions.

This document compiles the main ideas and recommendaƟons that emerged from the Workshop, both from the themaƟc sessions and the sub-group exercises, with the aim of strengthening strategies and policies to enhance the response capaciƟes of the Ministries of Labor in the region regarding the impact of AI on employment and the expansion of work on digital plaƞorms.

1. MAIN IDEAS AND GENERAL CONSIDERATIONS

Below is a consolidaƟon of the main ideas that emerged from the speakers’ presentaƟons in each session, from the intervenƟons during the dialogue spaces where Ministries and social actors shared their naƟonal experiences, and from the subgroup discussions. A later secƟon of this document consolidates the policy recommendaƟons idenƟfied throughout the workshop.

1.1. ArƟficial Intelligence, Employment and Skills

• ArƟficial Intelligence (AI) is evolving at an extraordinary pace, increasingly permeaƟng mulƟple spheres of life and economic sectors. A clear example is that the first version of ChatGPT appeared in November 2022 and is now in its fiŌh version, while new applicaƟons and plaƞorms are emerging daily. The speed and scope of AI present major challenges for regulaƟon and public policy, which risk being leŌ behind.

• In LaƟn America and the Caribbean, the adopƟon of AI is advancing rapidly. According to data

emerging daily. The speed and scope of AI present major challenges for regulaƟon and public policy, which risk being leŌ behind.

• In LaƟn America and the Caribbean, the adopƟon of AI is advancing rapidly. According to data shared by CENIA (Chile), the region recorded the fastest growth worldwide in generaƟve-AI penetraƟon and moneƟzaƟon last year, represenƟng 20% of the global market share. However, AI development and investment remain in early stages and far from other regions of the world; this is one of the major challenges. AI contributes only almost 1% of the region's GDP , while its global4

contribuƟon is 3.5% of GDP , and the region invests only 3.7% of global spending in arƟficial intelligence, as noted by ECLAC, revealing a wide margin for investment and development in AI in the region.

• The LaƟn American AI Index (ILIA for its acronym in Spanish), developed jointly by CENIA (Chile) and ECLAC, measures each country’s readiness regarding AI. Results for 2025 show three clusters of countries: Pioneers, which are countries whose ecosystems exhibit strengths in enabling factors, maturity, and governance: Chile, Brazil and Uruguay; Adopters, which have relaƟve strengths in at least one of those dimensions and include ArgenƟna, Colombia, Dominican Republic, Mexico, Peru, Costa Rica, Panama and Ecuador; and Explorer countries, which show significant gaps across all dimensions.

• It was recognized that AI can bring major benefits while also posing serious risks for humanity.

Among the warnings raised was that AI is not neutral and that generaƟve AI is the first technology capable of producing ideas and making decisions by itself, ciƟng Yuval Harari’s view of AI as an agent rather than merely a tool.

Among the warnings raised was that AI is not neutral and that generaƟve AI is the first technology capable of producing ideas and making decisions by itself, ciƟng Yuval Harari’s view of AI as an agent rather than merely a tool.

• Among the main challenges to AI development menƟoned were the lack of digital infrastructure, low connecƟvity, and high energy consumpƟon, as well as the shortage of local technological talent and the widening digital skills gap. The associated risks highlighted were job subsƟtuƟon and loss, threats to privacy and data protecƟon, algorithmic bias and discriminaƟon, and risks to physical and mental health.

• Among the main benefits of AI menƟoned were increased producƟvity levels, the emergence of new roles and occupaƟons, the ability for workers to focus on higher-value and higher-income acƟviƟes, and improved working condiƟons, acknowledging that the automaƟon of repeƟƟve or hazardous tasks can lead to safer working condiƟons. The opportunity that technology and AI represent in terms of labor inclusion was also menƟoned, enabling access for tradiƟonally disadvantaged populaƟon groups.

• There is broad consensus that the benefits, opportuniƟes, and risks of AI will depend on how this transformaƟon is managed. If technological transformaƟon is handled with strategic vision and inclusion, it can contribute to reducing inequaliƟes, promoƟng economic growth, and building more equitable socieƟes and economies across the Americas.

• There is also broad recogniƟon that inequality in the region is a major determinant of how the benefits and risks of AI are materializing and will conƟnue to materialize. It was highlighted that significant gaps are prevenƟng the region from fully leveraging the potenƟal of new technologies.

benefits and risks of AI are materializing and will conƟnue to materialize. It was highlighted that significant gaps are prevenƟng the region from fully leveraging the potenƟal of new technologies. Some revealing data shared by ECLAC show that although the region has tripled its internet connecƟvity levels in over the past 15 years, key challenges remain: broadband penetraƟon in LaƟn America is around 20%, while in the European Union it is close to 40%, while at the household level, there is a considerable penetraƟon gap between the poorest and richest households. At the business level, very significant progress has been made in connecƟvity, with nearly 90% of companies now connected to the internet; however, 96% of medium-sized and small businesses5

(where employment in the region is concentrated) lack an online presence, and almost 63% of them use the internet passively.

• Across the region, there is widespread interest and effort to measure how AI is impacƟng employment, although with clear differences among countries. Several countries — including ArgenƟna, Brazil, Canada, and Chile, among others — have already developed esƟmates using different criteria and analyƟcal methodologies to classify occupaƟons and assess their level of exposure to AI, the risk of displacement, as well as their potenƟal contribuƟon to producƟvity.

• To esƟmate how AI affects the world of work, it is important to recognize that occupaƟons are composed of mulƟple tasks, and that the impact will depend on how many of those tasks can be automated or replaced by arƟficial intelligence. This determines the impact not only on the quanƟty, but also on the quality of employment. The key quesƟon, therefore, is not “What is the

automated or replaced by arƟficial intelligence. This determines the impact not only on the quanƟty, but also on the quality of employment. The key quesƟon, therefore, is not “What is the impact of AI on employment?” but rather “What is the impact of AI on the tasks that make up jobs?” The methodology presented by the ILO analyzes all tasks within each of the 436 occupaƟons classified under the InternaƟonal Standard ClassificaƟon of OccupaƟons (ISCO) to measure their level of exposure to generaƟve AI and their risk of automaƟon.

• Below are some of the esƟmates presented by the ILO, OECD and ECLAC:

o For the Americas, the ILO esƟmates that 29% of employment is exposed to generaƟve AI (24% of male employment and 35% of female employment), although there are important differences within that 29%. For example, among professionals and technicians, certain tasks are exposed to AI while others are not, presenƟng opportuniƟes to increase producƟvity. Notably, within that 29% of employment exposed to generaƟve AI, 4.7% is at very high risk of automaƟon — equivalent to 22 million jobs, represenƟng 3% of male employment and 7% of female employment.

o ECLAC esƟmates that machine learning alone would affect around 35% of jobs in LaƟn America and the Caribbean, but that generaƟve AI raises this figure even further, to 44% of jobs. When disaggregaƟng by educaƟonal level, those most affected by machine learning are individuals with the lowest levels of educaƟon.

o AddiƟonally, the OECD esƟmates that 27% of employment across OECD countries faces high risk of automaƟon.

learning are individuals with the lowest levels of educaƟon.

o AddiƟonally, the OECD esƟmates that 27% of employment across OECD countries faces high risk of automaƟon.

o The ILO finds that more jobs are being 'transformed' rather than eliminated, while the OECD reports that, so far, there is no evidence that AI is causing net job losses, although some data are beginning to show impacts on specific groups. For example, new evidence from the United States indicates that generaƟve AI has led to a 13% decline in employment among early-career workers.

o Surveys conducted by the OECD seek to measure impacts on the quality of employment according to workers' percepƟons, among others: 79 % of workers who are using AI6

believe that it's improved their performance at work, and 63% report AI has improved enjoyment of their job.

• Studies conducted in various countries in the region corroborate the conclusions presented by internaƟonal organizaƟons: There is no evidence of massive job displacement, but there is evidence of a major transformaƟon of occupaƟons.

• AI has the potenƟal to boost producƟvity and generate value in occupaƟons where its role is complementary to human labor. AI does not replace human capabiliƟes, but rather amplifies them, contribuƟng to improve processes and the quality of results. To realize this potenƟal, it will be essenƟal to anƟcipate skills and competencies gaps and strengthen training, cerƟficaƟon, and reskilling systems so that workers can adapt to new market requirements. In this context, it is highly important for Ministries of Labor to develop labor foresight strategies and acƟons.

• In fact, it has been observed that, so far, AI is behaving more as a complement than a subsƟtute

reskilling systems so that workers can adapt to new market requirements. In this context, it is highly important for Ministries of Labor to develop labor foresight strategies and acƟons.

• In fact, it has been observed that, so far, AI is behaving more as a complement than a subsƟtute for exisƟng jobs, and that the risk of replacement in some sectors is accompanied by potenƟal employment growth in others. In ArgenƟna, for example, 71% of jobs would be enhanced by AI, while 22% are at risk of complete replacement (mainly administraƟve roles); for every replaceable job, there are three where AI would assist with tasks. 31% of Canadian workers are in highexposure roles (e.g., clerical support roles, telemarkeƟng) and highly suscepƟble to automaƟon; while 29% face a potenƟally posiƟve impact, with high exposure and low risk of displacement

(educaƟon, health and engineering sectors) and 40% would have a neutral (low exposure) effect. Meanwhile, CENIA (Chile) evaluated the level of task acceleraƟon resulƟng from AI within the 100 most common occupaƟons in the country and found that, overall, the equivalent of 12% of GDP would be added to the Chilean economy if generaƟve AI were fully adopted.

• AdministraƟve support jobs are at greater risk of automaƟon because they comprise a large number of tasks that can be automated by AI. The most affected workers will be those whose tasks are rouƟne, repeƟƟve, or data-driven, as well as those who lack mastery of AI-related skills.

Furthermore, generaƟve AI is impacƟng occupaƟons that were once considered "safe," such as those related to engineering or law, and can enhance and improve work in sectors such as

Furthermore, generaƟve AI is impacƟng occupaƟons that were once considered "safe," such as those related to engineering or law, and can enhance and improve work in sectors such as educaƟon and healthcare. Likewise, jobs related to science, technology, engineering, and mathemaƟcs (STEM) fields will conƟnue to see significant growth.

• There is a strong gender component to the impact of AI on employment, given that there are marked gender differences across sectors, with some sectors having a high female or male presence. For example, administraƟve support jobs are predominantly held by women and are at greater risk of being automated.

• There is a clear regional and global consensus on the central role of educaƟon and job training in this evolving landscape. These are recognized as essenƟal for maximizing the benefits of AI while miƟgaƟng the risks of job loss and displacement. Countries in the region are making progress7

in strengthening, reviewing, adjusƟng, and incorporaƟng innovaƟons into their training efforts, several of which are included in the "RecommendaƟons" secƟon later in the document.

• The skills gap, or the lack of the skills or competencies needed to harness the potenƟal of AI, affects both workers and businesses. In OECD surveys, more than two in five firms reported that lack of relevant skills in the workforce was a major barrier to adopƟng AI, surpassed only by its cost.

• SupporƟng micro, small, and medium-sized enterprises in developing skills and adopƟng generaƟve AI was idenƟfied as a significant need and opportunity. The enormous growth potenƟal this represents for these companies, as well as for the broader economy, was highlighted.

generaƟve AI was idenƟfied as a significant need and opportunity. The enormous growth potenƟal this represents for these companies, as well as for the broader economy, was highlighted.

• The region faces a digital skills gap, and exisƟng training strategies are insufficiently focused on addressing it. Training policies largely lack a direct emphasis on AI: only 33% of OECD countries provide publicly funded training in general AI literacy, and those who need it most—such as adult workers with lower levels of educaƟon—are oŌen the least likely to receive it.

• The gap with developed countries is considerable. Although the share of AI talent within the workforce in LaƟn America has doubled over the past eight years, no country in the region has yet reached the levels that countries in the Global North had at the beginning of the same period, according to ILIA results for 2024.

• Furthermore, the region faces a significant migraƟon of specialized talent, or “brain drain,” largely due to low investment and slow development of AI. Professionals working with fronƟer models, such as generaƟve AI and language models, are the ones migraƟng out of the region, according to CENIA data. Within this AI segment at the professional level, there is a pronounced gender gap: only one in four professionals in the field is a woman, reflecƟng paƩerns also observed in academia.

• Another challenge facing training is that it is concentrated among younger people. According to ECLAC, between 80 and 90% of training in the region is provided to people under 30 years of age.

This must be expanded to realize a life-cycle or lifelong learning approach.

• That said, most workers will not need specialized AI skills, but rather complementary and enabling skills. OECD surveys indicate that employers are placing greater value on skills that AI

This must be expanded to realize a life-cycle or lifelong learning approach.

• That said, most workers will not need specialized AI skills, but rather complementary and enabling skills. OECD surveys indicate that employers are placing greater value on skills that AI cannot replicate, such as social and emoƟonal skills, business processes, and management capabiliƟes. It was also noted during the Workshop that while not everyone needs to be an AI expert, everyone should know how to use AI effecƟvely.

• It was also menƟoned that high levels of informal employment make it more difficult to measure and analyze the impact of AI on employment in the region; furthermore, it may limit the acƟon and effecƟveness of public policy in this new scenario.8

• In general terms, the strategies implemented by the governments that were menƟoned include

the following:

o CreaƟon of intersectoral bodies or agencies to regulate the ethical and responsible use of AI, as well as to promote connecƟvity and digital transformaƟon processes. Some were menƟoned, such as Ecuador's AI CommiƩee, chaired by the Ministry of TelecommunicaƟons and InformaƟon Society (MINTEL, by its Spanish acronym), and Trinidad and Tobago's Ministry of Public AdministraƟon and AI, which is leading efforts to transform public service in the digital age, and Belize's NaƟonal Digital Agenda.

o AdaptaƟon of employment policies and improvements in training strategies and systems to develop digital skills. The iniƟaƟves menƟoned included the WeLearnTT program in Trinidad and Tobago, the Skills for Success Program in Canada, and “Talento Digital” (Digital talent) and “Hazlo con IA” (Do it with AI) in Chile. The laƩer defines personalized training pathways and offers e-learning courses with micro-content.

Digital” (Digital talent) and “Hazlo con IA” (Do it with AI) in Chile. The laƩer defines personalized training pathways and offers e-learning courses with micro-content.

o Improvements in public employment services and labor market informaƟon systems. MenƟon was made of “EmpleaPY” in Paraguay, which connects young people with quality jobs in high-demand sectors, as well as Jamaica’s “Integrated Labour Market InformaƟon System” (ILMIS), which is being modernized to process data from mulƟple sources and generate forward-looking analyses to inform policymaking. Chile’s modernizaƟon of their NaƟonal Employment Exchange (BNE, by its Spanish acronym) was also highlighted.

o Development of research, analysis, and foresight efforts to monitor labor market trends, idenƟfy the impact of AI on employment and skill requirements, and guide public policy. Canada's Future Skills Center was highlighted as a hub for applied research, innovaƟon, and training, as well as Chile’s NaƟonal Labor Foresight Strategy and Regional Labor Observatories.

• RepresentaƟves of workers and employers explicitly referred to AI governance, recognizing its importance in ensuring the ethical, responsible, and transparent use of AI and in achieving the effecƟve and pracƟcal applicaƟon of regulaƟon. They emphasized that the effecƟve parƟcipaƟon of workers and employers in AI governance must be guaranteed.

• The need to strengthen triparƟte insƟtuƟonalized social dialogue and collecƟve bargaining was recognized, and case studies were cited showing that when workers are involved in how AI systems are introduced in the workplace, these systems perform much beƩer both for the company

(through increased producƟvity) and for improved working condiƟons.

recognized, and case studies were cited showing that when workers are involved in how AI systems are introduced in the workplace, these systems perform much beƩer both for the company (through increased producƟvity) and for improved working condiƟons.

• ReflecƟons from COSATE: 1) The development of AI must take into account that the region is marked by poverty, informality, and inequality, where workers in many countries face violaƟons of labor rights; 2) triparƟte social dialogue, freedom of associaƟon, the right to strike, and collecƟve bargaining are fundamental; without them, progress on regulatory frameworks and effecƟve, sustainable public policies cannot be achieved; 3) to ensure AI complements rather than replaces9

jobs, educaƟonal and training policies are needed, grounded in social protecƟon that guarantees equal opportuniƟes, given that digital divides, insufficient income, unequal distribuƟon of care work, and discriminaƟon demonstrate unequal starƟng points. COSATE-TUCA shared the document “AI and Work in LaƟn America”, prepared by the TUCA Labor Observatory of the Americas.

• ReflecƟons from CEATAL: Three main opportuniƟes of AI were highlighted: 1) the possibility of facilitaƟng work and reducing the performance of basic tasks, allowing workers to focus on higher value-added acƟviƟes, 2) the need for a broad educaƟon and qualificaƟon policy to take advantage of the full potenƟal of AI, and 3) AI facilitates and provides a lot of informaƟon on working condiƟons, including health and safety at work, which allows for training and simulaƟons to be carried out for the benefit of workers and companies.

1.2. AI and emerging technologies for the benefit of Ministries of Labor

condiƟons, including health and safety at work, which allows for training and simulaƟons to be carried out for the benefit of workers and companies.

1.2. AI and emerging technologies for the benefit of Ministries of Labor

• The adopƟon of AI by Ministries of Labor across the region is progressing steadily and is projected to be a strategic tool for strengthening insƟtuƟonal management. As noted by the IDB, this adopƟon includes its use in employment services, job training, labor inspecƟon, occupaƟonal safety and health, social security, labor jusƟce, and labor dispute resoluƟon. It was highlighted that these technologies enable a shiŌ from reacƟve to proacƟve management, anƟcipaƟng risks, improving services, providing a broader overview of informaƟon, and designing evidence-based policies.

• It was highlighted that AI has the potenƟal to opƟmize rouƟne and administraƟve processes, generate greater insƟtuƟonal efficiency and effecƟveness, expand the scope of services, accelerate data processing, and reduce operaƟng costs. In Canada, the ESDC launched the EVA virtual assistant, based on generaƟve AI, to support officials in document draŌing, text ediƟng, preliminary research, and other tasks, under strict privacy and security standards. These tools have demonstrated a high level of adopƟon and have freed up resources to strengthen insƟtuƟonal innovaƟon.

• The need for AI to complement, rather than replace human judgment, was emphasized.

Experiences from other sectors presented by AudacIA, such as healthcare, show that involving professionals in the design and validaƟon of algorithms strengthens trust and improves the

insƟtuƟonal innovaƟon.

• The need for AI to complement, rather than replace human judgment, was emphasized.

Experiences from other sectors presented by AudacIA, such as healthcare, show that involving professionals in the design and validaƟon of algorithms strengthens trust and improves the adopƟon of soluƟons. In the field of labor, this principle is key to ensuring that technological modernizaƟon is aligned with people's rights and with the central mandate of the Ministries of Labor as guarantors of labor policy and the protecƟon of workers' rights.

• AI is increasingly being used to improve labor intermediaƟon, idenƟfy skill gaps early on, and facilitate alignment between labor market demands and training supply, as shown by the

following experiences:

o Canada has matching algorithms that link job seekers' profiles with vacancies based on skills and experience.10

o In Jamaica, the SPIRO project -funded by the World Bankis developing a naƟonal job portal with AI capabiliƟes that will enable explainable matching, idenƟfy skills gaps, and link the educaƟon system to industry needs.

o In Guatemala, there are plans to i