Beyond budgetary allocation, budgeting for AI curriculum needs real world intersections

Can Artificial Intelligence (AI) “think” like humans? Can AI predictions be trusted completely—Why or why not? What happens if Machine Learning models are trained with wrong data? Why should we not copy from AI? What should we do if AI gives the wrong answer? Can AI be fair and kind?

These are not just musings but critical paths of inquiry reinforcing the need for AI education in schools. By introducing AI, data science, and ethical digital practices early in schools we prepare our future workforce to be curious thinkers, problem solvers, innovators, informed digital citizens, and ethical AI practitioners.

This begs the question how are we navigating the pathway to introducing AI education in schools? The Ministry of Education, Government of India, has constituted a committee to draft a curriculum to introduce AI and computational thinking in schools from Class III. While this is welcome, introducing AI incrementally may deprive a whole generation of critical employable skills. We hope that the next Union Budget will include a provision for AI education in schools on a mission mode for AI education right up to Class XII—with speed and scale—to enable schools to match strides with industry expectations.

Building a digitally empowered India starts with empowering every child with future skills today. AI and disruptive technologies signal a 51% skill gap in India. The country needs more than one million AI-skilled engineers in two to three years, according to NASSCOM. The demand-supply mismatch will widen to 30% by 2028. Further, with the need for ‘Talent Acceleration’ in line with national priorities around IndiaAI Mission, India Semiconductor Mission, automation and precision technologies, the lack of a digitally-skilled workforce can become a bottleneck leading to significant opportunity cost. We are not discussing a subject. This is national strategy.

AI Education can be an important channel for educating our youth in real world scenarios—where students can learn that AI models learn from data and feedback to make predictions. But true intelligence comes from understanding its decisions and using them responsibly and fairly or from knowing that AI is not just about technology—it is about solving real problems responsibly, exploring new careers, and making a positive impact on society. It is understanding that ML is computers learning from data to help us solve problems. The real intelligence comes from how we use it responsibly or understanding that personalisation by generative AI can create “filter bubbles,” along one’s existing understanding or preferences—but collaborating with humans in a diverse dialogue will expand the boundaries of one’s knowledge.

In most learning situations, logic is not an active consideration during curriculum transaction—and yet an important curtain raiser on AI in education is the query ‘What makes logic important for AI’? Thus, sceptics who may be prone to trash the idea of integrating AI in school curriculum as a fancy notion, may be persuaded to expand their belief systems as AI literacy can accelerate the building blocks of cognitive skills for developing unique competencies like critical thinking, complex problem-solving, collaboration, communication, empathy, ethical reasoning, adaptability systems thinking. Provided the pedagogy can go beyond boring theoretical definitions to embrace experiential inquiry-based learning strategies such as discussion-led exploration of real uses, guided use of safe online tools, activity-based appreciation of industrial applications.

Drafting a policy on AI education for K-12 students would not only need to be supported by Budgetary outlay but budgeting the curriculum would need real world intersections requiring adequate educator training, infrastructure, tools and software. Contemporary curriculum coverage will need play-based, gamified tools, real-world projects, and exposure to global AI tools and trends. There must be careful selection of real-world connections with experiential learning activities to allow: Early applied AI thinking (understanding ChatGPT’s responses); Debunking myths (What happens when AI makes a wrong/unfair decision?); Digital citizenship mindfulness (internet basics, cyber hygiene, responsible sharing, safe browsing); Ethical considerations (fairness, privacy, and empathy in machines, bias in facial recognition or hiring algorithms); Pre Coding exposure (Pattern recognition, algorithmic thinking without code); Data and coding fluency (Training a basic sentiment model); college and career readiness (AI and society – AI and careers; impact of automation in industry).

For educators, AI’s use for predictive checkpoints of learning can extend their reach into the hidden learning blocks that impede classroom progress and learning outcomes.

Thus, engaging with questions such as what counts as valid and viable AI education, and who gets to decide would need a “Responsible AI in Education” framework, while non-binding guidelines not prescribing specific technologies or content would need to consider accountability, and pedagogical soundness. There must also be a “Quality AI Seal” for programmes that demonstrate pedagogical effectiveness, cultural sensitivity, industrial relevance and commitment to data protection and privacy.

Views are personal. The author is a Co-Founder and Managing Director, Primus Partners