Generation AI: fears of ‘social divide’ unless all children learn computing skills
Education leaders say AI literacy should become a core school subject to stop a widening gap between those who understand AI and those who don’t. Declining computing GCSE entries and rapid AI adoption underscore the urgency.
In a Cambridge primary school, 10‑year‑old Joseph taught his own AI model to tell the difference between a drawing of an apple and a sketch of a smile. When the system confused a fruit for a face, he quickly retrained it, demonstrating a level of intuition about machine learning that many adults still lack. Joseph and his classmates at St Paul’s C of E primary school are part of a new generation that grows up surrounded by artificial intelligence, just as previous generations grew up with flight or social media.
Why AI literacy is being called a societal necessity
Philip Colligan, chief executive of the Raspberry Pi Foundation, warns that without universal AI education a "big split" could emerge between citizens who can understand and control automated decisions and those who cannot. He argues that AI literacy should sit on equal footing with reading and writing because algorithms now influence housing allocations, welfare benefits, health diagnoses, criminal‑justice outcomes and financial services.
Simon Peyton Jones, a leading computer scientist who helped design the UK’s national computing curriculum in 2014, echoes this concern. He calls for a new digital‑literacy qualification that teaches students to view AI as a transparent tool rather than a mysterious black box. "If you don’t know how the magic works, you are terribly disabled," he told the Guardian.
Falling interest in traditional computing amid AI hype
Data from the UK exam regulator shows a sharp decline in GCSE entries for computing, with only about 2,025 students sitting the exam in 2025—a fraction of those choosing history, biology, chemistry or physics. At the same time, a Ipsos poll recorded a 78% rise in AI system usage across the country over the previous year.
Big AI firms have contributed to the perception that coding may become obsolete. Anthropic CEO Dario Amodei claimed that 90% of his company’s own code is now generated by its Claude model. The phrase "vibe coding" entered tech jargon in 2025, describing the idea that natural‑language prompts could replace traditional programming. Even political figures have weighed in; former Labour leader Keir Starmer suggested in 2023 that teaching children to code might be outdated in an AI‑driven future.
What educators and charities are doing
The Raspberry Pi Foundation, which supplies low‑cost hardware to schools worldwide, is actively lobbying governments to embed AI concepts into curricula. Colligan says teachers are receiving mixed messages: on one hand, they hear that coding is no longer essential; on the other, they see a growing need for students to understand how AI makes decisions that affect daily life.
In Joseph’s coding club, children learn the basics of data sets, model training and bias detection. The experience has already shaped Joseph’s perspective: he believes AI can be beneficial but worries that unchecked errors could harm society. "If lots of people believe it when it’s wrong, it will have a bad impact," he said, adding that he wants to stay in charge of the technology rather than let it dictate outcomes.
Potential consequences of an AI‑illiterate generation
Colligan warns that socioeconomic disparities could widen. Children in well‑resourced schools that can afford AI‑focused lessons will likely become more empowered citizens, able to question automated decisions and advocate for their rights. Conversely, pupils in under‑funded areas may become passive recipients of algorithmic outcomes, with little ability to challenge them.
Former deputy prime minister Nick Clegg, now an AI investor, recently predicted that society will shift from "staring at the internet" to "living in the internet." Without a solid grounding in how those digital environments operate, many risk losing agency over personal and public choices.
What comes next for AI education
Policymakers are under pressure to revise the national curriculum to include a dedicated AI literacy component. Proposals include mandatory modules on data ethics, model interpretability and practical hands‑on projects similar to Joseph’s classroom exercises. Industry groups are also being asked to fund teacher training and provide open‑source tools that demystify AI for young learners.
Until such reforms are codified, the debate will continue in classrooms, boardrooms and parliament. The stakes are high: as AI systems become ever more embedded in public services, the ability to understand, critique and direct those systems may determine whether future citizens are active participants or silent subjects of automated decision‑making.
Why it matters
AI literacy will determine whether future generations can engage with, challenge and benefit from automated systems that increasingly shape public life.
Key points
- AI literacy is being framed as a fundamental skill comparable to reading and writing
- UK computing GCSE entries have fallen sharply while AI usage has surged
- Industry leaders claim coding may become redundant, fueling debate about curriculum relevance
- Socio‑economic gaps could widen if AI education is unevenly distributed
- Policymakers are being urged to add dedicated AI modules to the national curriculum
Frequently asked questions
Why do experts compare AI literacy to reading and writing?
Reading and writing are basic tools for participation in society; similarly, understanding AI is becoming essential for navigating decisions that affect housing, health, finance and justice.
What is the Raspberry Pi Foundation’s role in AI education?
The charity supplies affordable hardware to schools and campaigns for AI literacy to be taught universally, arguing that lack of knowledge could create a dangerous social divide.
How are AI companies influencing the perception of coding education?
Companies like Anthropic promote the idea that AI can write most code, leading some to believe traditional programming skills are becoming obsolete.
What risks do children face if they lack AI literacy?
They may be unable to question or contest automated decisions, leaving them vulnerable to unfair outcomes in services such as welfare, healthcare and criminal justice.
