AI Can Generate Code. Is That a Threat to Computer Science Education? (www.edweek.org)

High school students and teachers are grappling with a new reality: generative AI can write code, and headlines about layoffs and CEOs claiming coding will be unnecessary are stoking anxiety about future tech careers. Education leaders and organizations push back, arguing that coding and foundational computer-science skills remain crucial because humans must guide, evaluate and control AI systems. The Raspberry Pi Foundation outlines five reasons to keep coding in curricula: humans needed to direct AI outputs; coding builds computational thinking; it expands future opportunities as AI reaches other industries; it provides digital agency to contest automated decisions; and it enables learners to shape technologies rather than be shaped by them. Practically, that means evolving how computer science is taught: integrate probabilistic, data-driven models and AI literacy, teach students to use and critically evaluate models (including societal and ethical impacts), and shift assignments and assessments so AI is a tool not a shortcut. Teachers are already using large language models for brainstorming and debugging—acting like a “rubber duck” to surface student reasoning—but classrooms will need new lesson designs, professional development, and funding to scale this. Policy attention to AI in K–12 is rising, but experts stress sustained investment to ensure students gain the technical fluency (data literacy, system design, ethics) needed to steward AI responsibly.