- AI is already in classrooms—and being used creatively to support learning. From literacy tools with speech recognition to platforms that personalize instruction, educators are using AI to strengthen teaching, not replace it.
- Inclusion and equity are emerging as central drivers of AI innovation. A significant share of initiatives focus on supporting students with disabilities, diverse learning needs, and vulnerable trajectories—showing promising potential for reducing gaps.
- Responsible, scalable implementation remains the region’s biggest challenge. Most initiatives acknowledge ethical risks and the need for evidence, but few have rigorous evaluations or robust safeguards, underscoring the importance of governance, teacher training, and data systems.
Across Latin America and the Caribbean, teachers and school leaders are asking the same question: How can we use artificial intelligence to help every student learn better?
For the first time, we now have a regional snapshot of what is actually happening inside classrooms in the region—not theories, not predictions, but real tools already in motion. That’s the spirit behind the Inter-American Development Bank’s first call for artificial intelligence solutions to transform education called “AI Here!” (¡IA Presente!). This call for solutions in partnership with Ceibal and implemented by Socialab aimed to surface real-world AI applications already operating in schools, and to begin moving from promise to proof.
In response, 193 initiatives across 22 countries pitched their use of AI in schools — from adaptive student learning tools to intelligent systems for school management. The main message: AI is no longer just “coming”—it’s already here. What matters now is how we make it work for all students.
In April 2025, the IDB released AI and Education: Building the Future Through Digital Transformation, a report that examines the role of artificial intelligence through the lens of what we already know from decades of digital education. The evidence is clear: while technology can improve learning and expand opportunities, it only works when it is meaningfully integrated into teaching, adapted to local contexts, and deployed equitably. AI is no exception—its value depends far more on how we use it than on the technology itself.
The report highlights three areas where AI can have the greatest impact:
- Maximizing learning in the classroom
- Ensuring inclusive and continuous educational trajectories
- Enabling more efficient school management.
Given the limited evidence on real uses and impacts of AI in schools across Latin America and the Caribbean , the “AI Here!” call was launched with a clear purpose:
- To map AI solutions already being used in real classrooms.
- To identify what is working, for whom, and under what conditions.
- To highlight obstacles—especially the ones that show up when scaling.
- To inform public policy decisions by making visible the link between AI practice and impact.
The call invited submissions from startups, NGOs, universities, public agencies, or hybrid partnerships—but importantly, only tools already deployed in schools were eligible. That ensures we focus on practice, not prototypes.
- AI is already being used to boost classroom learning: More than half of the submissions (57%) focused on learning inside the classroom — adaptive platforms, literacy apps using speech recognition, gamified pods for math/science, teacher-copilot tools helping lesson planning and feedback. This shows a region actively seeking to use AI not as add-on tech but as pedagogical amplifier.
- Inclusion is becoming a real priority: More than 25% of the solutions targeted inclusion and learning continuity — students with disabilities, socio-emotional needs, access issues, remote/underserved areas. In a region where large achievement and access gaps persist, the fact that more than a quarter of submissions target inclusion signals a positive turn.
- School-management and system-level solutions remain under-served: Only ~16% of initiatives addressed management, monitoring, system-level decisions (budgets, procurement, infrastructure, dashboards). That’s the frontier: there’s tremendous potential for AI to lighten administrative burdens, free up teacher time, and help Ministries make evidence-based decisions. But it’s still early.
- The region is not just importing tools — it’s innovating combinations of AI: Technically speaking: 59 % of initiatives used generative AI, 27% used language models, 24% natural-language processing, 12% image/video recognition, 12 % speech/voice recognition. About 48% combined multiple AI modalities. In short: Latin America is building hybrid, context-sensitive AI systems—not simply copying models designed elsewhere.
From this large pool, a selection committee lead by IDB and Ceibal, identified four initiatives that offer strong inspiration for policymakers and community actors:
| Empujón Educativo (Argentina): This initiative uses ethical AI, neuroscience, and behavioural insights to craft a dynamic “educational story” for each student — personalised pathways that adapt over time while keeping high expectations. Learn more in this video. |
| Aprender Haciendo with Amira Learning (Costa Rica): AI listens to children reading in Spanish and English, analyses fluency and comprehension, and gives real-time feedback. Key point: literacy tools grounded in the regional languages and classroom environments. Learn more in this video. |
| Rio de Janeiro’s Education Secretariat (Brazil): Here AI automates validation of school invoices (DANFEs) — a seemingly mundane but hugely impactful task. The result? Principals and district teams spend less time on paperwork and more on learning improvement. Learn more in this video. |
| Special correspondent - Vínculo® (Brazil): A platform for inclusive education built on an AI engine trained for students with disabilities; helps teachers personalise plans, track progress, and create inclusive experiences for diverse learners. Learn more in this video. |
These four paint a picture of the “sweet spot” for AI in education: pedagogy + context + scalability.
1. Ethics and Responsible Use: AI introduces powerful possibilities — but also risks. In the call:
- ~57% of submissions recognized the risk of bias in algorithms.
- Yet only ~27% described concrete mitigation strategies, and about 24% referenced principles like transparency or explainability.
That means: responsible AI still needs to move from agenda to implementation. For policymakers: it’s not enough to procure “AI tools”—you need governance frameworks, ethical audits, data-protection protocols, teacher training, and monitoring.
2. Evidence and Impact: While ~70% of submissions reported some evaluation, only a minority had rigorous impact studies. For education ministries and funders, that means scaling prematurely risks investing in tools without well-understood results. There’s a real need to strengthen the evaluation culture: run randomized trials, longitudinal follow-up, cost-effectiveness analyses, and open evidence dissemination.
AI will not magically fix the learning crisis. But this unprecedented mapping of 193 initiatives shows that Latin America and the Caribbean are already turning the corner: from technology talk to tools in action.
As policymakers, education system leaders, community advocates, and teacher-educators, the priority now is to scale up what works, listen to what doesn’t, and stay anchored in equity. AI should serve every student, in every school, regardless of zip code, language, or learning need. Let’s ensure that this is an education revolution for everyone — not just the well-connected.
Download the IDB report on AI and Education in the region, follow us on social media, and share your ideas - we’re just getting started.