India’s AI Education Push Faces a Crucial Test: Empowerment or a New Digital Divide?

The new curriculum seeks to develop problem-solving abilities, creativity, ethical decision-making, and responsible engagement with technology. For students in Classes 6–8, it includes introductory AI concepts, awareness of algorithmic bias, digital citizenship, and interdisciplinary projects designed to connect technology with real-world challenges.

The initiative arrives at a critical moment. Artificial intelligence is rapidly transforming education, employment, communication, and access to information. Properly deployed, AI has the potential to help students master difficult subjects, translate content across languages, generate educational materials, and provide personalized learning support. For children in remote or underserved communities, AI-powered tools could expand access to knowledge in ways previously unimaginable. Students who lack access to specialized teachers or advanced educational resources may gain individualized explanations and learning assistance that even many well-funded institutions struggle to provide.

Yet alongside this promise lies a profound concern: artificial intelligence could become a new source of social stratification.

This does not imply that technology will recreate India’s traditional caste hierarchy in a literal sense. Rather, AI risks creating a new division between those who design, control, and understand intelligent systems, and those who are merely governed by them. In such a future, power would be concentrated among the technologically literate, while others would find themselves increasingly dependent on systems they neither control nor fully comprehend.

The contrast could emerge most clearly in education itself. Students in affluent English-medium schools are likely to benefit from high-speed internet, personal devices, trained educators, and access to advanced commercial AI platforms. They will learn how to evaluate AI-generated information, challenge inaccurate outputs, craft effective prompts, and build applications of their own. By contrast, students in poorly resourced schools may encounter AI only occasionally, through shared hardware or simplified tools designed by others. One group will learn to direct intelligent systems. The other may simply learn to accept their decisions.

Emerging research suggests that such risks are not merely theoretical. Preliminary Indian studies have indicated that some AI systems produced simpler educational explanations for profiles associated with marginalized castes or students educated in Hindi or regional-language environments, apparently treating those characteristics as signals of lower ability. Another preprint examining Indian labor-market data found that Scheduled Caste and Scheduled Tribe graduates were less likely than upper-caste graduates in the same districts to work in occupations positioned to benefit from generative AI technologies.

While these findings remain subject to further scholarly scrutiny, they point to a broader reality. Artificial intelligence does not enter society as a neutral force operating in a vacuum. It is trained on existing data, generated within existing institutions, and deployed within existing social structures. As a result, it can absorb and potentially amplify inequalities that already exist.

The challenge extends beyond unequal access. It also involves the concentration of power.

Advanced AI systems require enormous volumes of data, sophisticated computing infrastructure, specialized chips, significant energy resources, technical expertise, and substantial financial investment. Consequently, the most capable systems are controlled by a relatively small number of corporations and governments. These actors possess considerable influence over what information becomes visible, which languages receive priority, what behaviors are rewarded, and who is regarded as employable, trustworthy, or worthy of access to services.

This concern is addressed directly in Magnifica Humanitas, the encyclical of Pope Leo XIV. The document does not reject technological innovation. On the contrary, it acknowledges technology’s potential to educate, heal, connect communities, and protect creation. However, it argues that technology is never neutral in practice because it reflects the intentions and priorities of those who design, fund, regulate, and deploy it. Artificial intelligence can support participation and justice, but it can also deepen inequality, exclusion, and social control.

For India, this warning carries particular relevance. Historically, caste-based disadvantage has often been reinforced through unequal access to knowledge, occupations, economic resources, and social visibility. AI could create similar patterns in less visible forms. Algorithms may shape opportunities and outcomes without openly revealing the assumptions embedded within their design.

The consequences are already evident in many areas where algorithms influence decision-making. An AI system may affect employment prospects, creditworthiness, or access to services while offering little transparency regarding how those conclusions were reached. Data can appear objective even when it reflects the biases of the societies from which it was collected.

There are also wider concerns regarding the misuse of artificial intelligence. AI systems can generate convincing misinformation, automate surveillance, manipulate behavior, strengthen authoritarian control, and facilitate cyberattacks. They may also reduce barriers to potentially dangerous biological or chemical knowledge. The use of autonomous weapons raises additional ethical questions by increasing the distance between human decision-makers and the consequences of lethal force.

The 2026 International AI Safety Report acknowledges that some AI-related harms are already occurring, while others remain uncertain but potentially severe. It also notes that developers cannot always predict the behavior of increasingly capable systems. Some leading AI scientists and technology executives have argued that the possibility of advanced AI posing an existential threat to humanity deserves serious attention alongside risks such as pandemics and nuclear conflict. Although such scenarios remain debated rather than universally accepted, the scale of potential harm makes caution a reasonable response.

These realities suggest that AI education must go beyond technical training and workforce preparation. Students should be encouraged to ask critical questions: Who created this system? Whose data was used to train it? Who benefits financially from its deployment? Which communities might be misrepresented? What human capacities could be diminished through automation? And should every technologically feasible application be permitted?

According to Magnifica Humanitas, learning to use AI includes learning when it should not be used. Genuine education depends on patience, reflection, dialogue, and engagement with reality. While AI can provide instant answers, schools must ensure that students continue to develop the habits of critical thinking, deep reading, and independent judgment.

Catholic education in India is uniquely positioned to contribute to this effort. AI literacy can become an extension of the Church’s longstanding educational mission. Catholic institutions can help ensure that students from Dalit, tribal, rural, and economically disadvantaged backgrounds gain meaningful access to AI technologies through shared laboratories, affordable devices, scholarships, and community learning centers.

They can also help broaden linguistic inclusion. AI education should not privilege English at the expense of regional and tribal languages. Educational resources and datasets developed in vernacular languages can help ensure that intelligence and potential are not measured by fluency in English.

Teacher formation is equally important. Educators must understand not only how AI functions but also where it fails, how bias emerges, and how students may misuse these tools. Teachers should remain mentors in wisdom and discernment rather than becoming mere supervisors of machine-generated work.

Ethics must also occupy a central place in AI education. Courses should address privacy, surveillance, deepfakes, caste and gender bias, job displacement, environmental impacts, autonomous weapons, and the concentration of corporate power. Students need exposure not only to technical skills but also to philosophy, theology, social science, human rights, and civic responsibility.

Ultimately, schools must preserve what technology cannot replace: friendship, conscience, compassion, moral courage, embodiment, silence, and the search for meaning. The purpose of education is not to create children who function like efficient machines. It is to form human beings capable of exercising power responsibly.

AI does not have to become India’s new caste system. It can serve as a tool for inclusion, opportunity, and shared learning. But that outcome is not inevitable. It will require deliberate choices about access, ethics, governance, and education.

The central question facing India is not whether its children will use artificial intelligence. They almost certainly will. The more important question is whether they will remain free, morally responsible, and fully human while doing so.

• Raju Hasmukh with files from Global Catholic