UK National Education Union poll: 66% of secondary school teachers in England have observed their pupils' critical thinking skills decline due to AI usage
Two-thirds of secondary school teachers report a decline in core abilities such as writing and problem-solving
Context & Ripple Effects
This extends a UK education arc in which student AI use has become commonplace: a prior survey found AI-assisted study use among UK undergraduates rose sharply, while analysis of submitted work had already identified material levels of AI assistance across a large student-paper corpus.
The new teacher-reported signal shifts the concern from authorship and academic integrity to whether routine reliance is displacing practice in writing and problem-solving. It also arrives after a recorded rise in proven AI-related academic-integrity cases in UK education, making assessment design a central operational issue rather than a marginal misconduct question.
First-order effects
- Secondary teachers and school leaders face a clearer reported skills-risk signal: pupils’ AI use is being associated by respondents with weaker writing, problem-solving, and critical-thinking practice.
- The poll strengthens the case for schools to distinguish AI-supported work from independently produced work when evaluating learning, even though it does not by itself establish causation.
Second-order effects
- Assessment providers and schools are likely to face pressure to use more supervised, oral, iterative, or process-based assessment, because take-home outputs become less reliable evidence of an individual pupil’s capabilities.
- AI vendors serving education face a sharper product-design and adoption test: tools positioned as learning supports will be judged against whether they preserve active practice, not merely whether they speed completion.
Third-order effects
- If similar findings persist across cohorts and settings, education systems may move from treating generative AI chiefly as a cheating-detection problem to redesigning curricula and assessment around demonstrable reasoning and authorship.
- The longer-term divide may be between institutions able to integrate AI with structured practice and verification, and those where convenience-led use erodes the feedback loop that builds foundational skills.
The trend: Generative AI’s spread through education is turning the debate from access and misconduct toward how institutions can preserve independent cognition while adopting productivity tools.