Artificial intelligence is advancing rapidly, reshaping how we work, learn, and interact. Yet in societies with high poverty rates and fragile education systems, this progress doesn’t reach everyone equally. Many people are left on the sidelines, fearful or feeling useless in the face of a technology that seems out of reach. If we don’t act, the gap between those who can use these tools and those who barely understand them will keep widening.

The challenge isn’t just teaching people to use AI, but helping them do it with critical awareness. In communities without strong educational support, people may see these systems as “infallible,” blurring the line between their own reasoning and a machine’s output. That can lead to bias, loss of judgment, anxiety, psychosis, and even suicidal thoughts. Dissociation and other mental health problems are already emerging in places where AI is consumed without guidance or filters.

That’s why balancing the conversation is key. Pushing AI adoption without tools for understanding is as risky as banning it outright—both leave people powerless. What’s needed is education: critical digital literacy, workshops that explain how systems work, their limits, and why they never replace human judgment. Showing that AI is a support, not an authority, helps people see it as part of everyday life rather than a threat.

There are initiatives already working along these lines:

• CSP of West Alabama – Digital Navigator Program: libraries turned into digital learning hubs where people with no prior experience learn to use devices and basic online services, gaining confidence and independence.

• Digital Poverty Alliance (UK): explores how generative AI can support education in areas affected by digital poverty, training teachers and tailoring technology for students who might otherwise fall behind.

• Tutor CoPilot (US): an AI system that supports tutors working with underserved students, improving learning outcomes for those most at risk.

• Morocco school dropout project: uses predictive AI models to identify vulnerable students and provide early interventions.

• Recode (Brazil): creates digital inclusion centers in rural, Indigenous, or prison communities, offering free access to computers, training, and support.

• Fundación Proacceso (Mexico): a network of learning centers that provides internet access, computers, and courses in low-income areas.

What these programs share is simple but vital: they don’t just provide technology, they offer human guidance and training, adapting content to local cultures and needs. This makes AI a shared resource rather than a privilege.

Technological progress doesn’t have to push people into exclusion or emotional harm. To prevent that, governments, schools, businesses, and social organizations need to provide programs that bring those outside the digital radar into the conversation: basic training, discussions on ethics and mental health, clear guidance for spotting false or biased content.

The goal is for everyone—regardless of background—to use AI with confidence and discernment while maintaining their identity and human value. Technology is powerful, but our capacity to reason, feel, and choose remains central. Teaching people how to balance both dimensions is the safest way to ensure AI doesn’t leave invisible scars but instead expands opportunity for all.

Governments are beginning to take action, but their approaches vary. The European Union and some Latin American countries have introduced digital literacy frameworks and laws such as the AI Act, along with inclusion programs that use libraries or community centers to teach basic digital skills. Still, many initiatives stall at the pilot stage or depend on donations, with no stable funding or coordination across agencies.

Efforts also tend to focus on technical skills, overlooking mental health, bias, or the cultural impact of algorithms. In rural areas or communities with native languages, there’s often little tailored content or sustained support.

What works best is cooperation between governments, NGOs, universities, and local networks: combining funding, teaching expertise, and proximity to those most in need. This approach turns AI into a shared opportunity, rather than a force that deepens inequality or erodes social trust.