July 30, 2026, (Inside AI) — Mexican entrepreneurs and educators are deploying artificial intelligence to solve deeply local problems, from chaotic recycling systems to under-resourced classrooms. The initiatives highlight a grassroots AI movement that is reshaping how communities across the country tackle infrastructure and education gaps.
In Querétaro, high school teacher Alma Trejo has built customized AI agents for her 56 students at Conalep, Mexico’s technical education system. Using Microsoft 365 Copilot and other tools, she created academic agents that explain concepts, propose exercises, and provide individualized feedback. There is also a programming tutor for computational thinking and another agent for digital culture and cybersecurity.
Trejo’s pivot began when AI started appearing in student assignments. Rather than resist, she pursued free Microsoft training courses, starting with AI for teachers, Copilot, and prompt writing. She later completed a Microsoft course on the fundamentals of generative AI and cybersecurity training. Now, she is pursuing a master’s degree in AI through a Conalep scholarship.
“We cannot allow ourselves not to apply AI tools in education,” Trejo says. “Otherwise, we are not preparing them for the present they need, or for the future.”
The impact was rapid. Students who once submitted work they couldn’t explain began discussing their reasoning. Quieter students grew more engaged, using AI tools to review concepts and build confidence. Trejo, a mother of three with a background in mathematics and sciences, sees AI as an ally in making learning personal and accessible, especially in a community with limited infrastructure and a small computer lab.
Meanwhile, in another part of Mexico, entrepreneur Ortiz is applying AI to recycling. Before participating in a training program, he had only heard about AI through media. “It opened the door to using artificial intelligence tools and to understanding what we can do with them,” Ortiz says. He is now seeking funding for an app that combines technology with community action to help residents participate in a more organized and sustainable recycling system.
These efforts reflect a broader trend: AI skill-building programs are enabling Mexicans to address local challenges without waiting for top-down solutions. Microsoft’s free courses, which Trejo utilized, are part of a global push to democratize AI education. However, critics note that such corporate-led initiatives can create dependencies on specific platforms, and the long-term sustainability of these projects remains uncertain.
Mexico’s AI adoption faces structural hurdles. According to the Organisation for Economic Co-operation and Development (OECD), the country lags in digital infrastructure and STEM graduates compared to peers. Yet, the grassroots nature of these projects suggests a bottom-up model that could bypass institutional bottlenecks. A 2023 OECD report on digital education highlights the importance of teacher-led innovation in bridging the digital divide, a principle Trejo embodies.
AI as a bridge, not a crutch
Trejo’s approach challenges the narrative that AI in education leads to cheating. By integrating tools like Copilot into her pedagogy, she shifts the focus from rote answers to critical thinking. Her agents do not replace instruction but extend it, offering support beyond classroom hours. This model could be replicated in other under-resourced settings, though it requires teacher training and reliable internet, both scarce in many Mexican regions.
Ortiz’s recycling app similarly leverages AI for civic engagement. While details of the technology are sparse, the concept aligns with global smart-city trends where machine learning optimizes waste management. A 2021 study in Waste Management found that AI-driven recycling systems can increase efficiency by up to 30%, but community participation remains the key variable. Ortiz’s focus on combining technology with community action may be its strongest asset.
The platform dependency question
Both initiatives rely heavily on Microsoft’s ecosystem. While this provides immediate access to powerful tools, it raises concerns about vendor lock-in and data privacy. Mexico lacks comprehensive AI regulations, leaving users to navigate terms of service without local safeguards. As these projects scale, they may need open-source alternatives or government-backed infrastructure to ensure long-term viability.
For now, the stories of Trejo and Ortiz illustrate a pragmatic, problem-first approach to AI. They are not chasing hype but using available tools to fill gaps in education and public services. The next test will be whether such efforts can attract sustained investment and policy support, or if they remain isolated bright spots in a challenging landscape.
