Skills, Trust & Consumer Adoption

Public backs AI that helps pupils learn but rejects it doing their work

Friday, 15 August 2025 3:51PM UTC

New Opinium polling for The Observer shows strong public support for using AI in schools as an assistive tool—for research, lesson planning and routine admin—while drawing a firm line against AI completing pupils’ essays or teachers using it for assessment; the findings point to urgent needs for national standards, teacher training, assessment redesign and stronger vendor accountability.

Opinium’s new polling, carried out for The Observer, reinforces a clear public instinct about artificial intelligence in schools: use it to help pupils learn, not to do their learning for them. The survey of 2,050 UK adults, conducted between 6 and 8 August 2025, found a striking split in popular sentiment. Half of respondents (50%) think pupils should be allowed to use AI for research, but three quarters (73%) oppose pupils using AI to complete essays or homework; just 17% say that should be permitted. Support for explicit teaching about AI is strong — two thirds (66%) want schools to educate pupils about emerging technologies — but people draw a firm line between “assist” and “replace” when it comes to classroom use.

That central judgment — AI as an aid, not a substitute for human responsibility — runs through a series of recent studies and sector reports. Opinium’s broader analysis of public responses to the government’s AI Opportunities Action Plan shows acceptance tends to track task risk: routine administrative automation attracts far more public support than AI-led planning or high‑stakes decisions. The same analysis flags the conditions the public wants before entrusting AI with sensitive roles: demonstrable accuracy, robust regulation, clear lines of accountability and strong data security.

The ambivalence about teachers’ use of AI is notable. The Observer poll found opinion narrowly split over teachers using AI for lesson planning (41% in favour, 42% opposed) and slightly more resistant when it comes to administrative paperwork (38% in favour, 43% opposed). Opposition hardens around assessment: 58% of respondents oppose teachers using AI to mark essays or homework, suggesting the public views marking as an inherently human responsibility tied to professional judgment and accountability.

Those public instincts align with evidence from schools and leaders. Browne Jacobson’s School Leaders Survey reports that around half of leaders say their schools use AI tools to some extent and about one in five use them regularly. Many leaders report efficiency gains in lesson preparation and in tailoring material to pupils, but also flag shortfalls in staff expertise and concerns about plagiarism, safeguarding and legal compliance. Similarly, Cambridge University Press & Assessment’s nationally representative polling of over 2,200 adults found broad backing for AI to assist with routine tasks but strong distrust of its use in assessment, with younger respondents consistently more receptive than older ones.

Practice on the ground is moving faster than policy in some respects. The Department for Education’s omnibus surveys show that many secondary pupils already use generative AI in lessons and for homework, underlining that classroom familiarity with these tools has risen sharply. Internationally, educators report similar trends: a survey of US K‑12 educators and administrators found rising optimism about AI’s potential to reduce workload and help with drafting communications and lesson materials, even as concerns about cheating, privacy and the limited reliability of detection tools persist.

Taken together, the evidence points to a pragmatic middle way. There is appetite for using AI to lift routine burdens from teachers and to equip pupils with the skills they will need in an AI-shaped economy — but only if that adoption is governed, resourced and explained. Several recurring policy priorities emerge from the data and sector reports:

Clear national standards and guidance. The public and school leaders want a consistent framework that sets out what AI may and may not be used for in classrooms, how assessment should be handled, and minimum data‑protection and transparency requirements.
Investment in teacher training and capacity‑building. Where schools report benefits — in lesson planning, personalised learning and efficiency — those gains are often tied to leaders who have invested in staff training. Government funding for continuing professional development would help spread good practice and reduce uneven adoption between better‑resourced and less‑resourced schools.
Assessment redesign and academic‑integrity measures. Given the strong public resistance to AI marking and the widespread concern about cheating, policymakers and exam boards should accelerate work on assessment formats and integrity tools that remain meaningful in an AI era, while ensuring detection technologies and sanctions are proportionate and fair.
Vendor accountability, transparency and data security. The public’s conditions for trust — accuracy, regulation, accountability and data security — imply procurement standards that require vendors to publish model capabilities, limitations, training data provenance where feasible, and clear contractual protection for pupil data.
Curricular literacy for pupils and parents. With many pupils already using generative AI, the surveys point to an urgent need for schools to teach responsible use, digital literacy and the ethical dimensions of AI, and for parents to be brought into the conversation so home use is consistent with school expectations.
Equity of access. To realise the promise of personalised learning and reduced teacher workload, policymakers must tackle disparities in access to devices, connectivity and staff expertise so AI does not widen existing attainment gaps.

These measures would also help the UK position itself as a leader in responsible educational innovation. The public’s cautious optimism — supportive of education about AI and selective practical uses but wary of automation that erodes professional judgement — gives policymakers a clear mandate: foster the positive opportunities while putting firm guardrails around riskier applications. As Opinium’s findings make plain, that balance is both politically sensible and pedagogically defensible.

There are practical steps schools and local authorities can start taking now. School leaders who have already piloted AI tools advise focusing on small, well‑scoped pilots that measure impact on teacher workload and pupil outcomes; investing in staff CPD; publishing transparent local policies on permitted and prohibited uses; and engaging parents and governors in policy design. Exam boards and inspection bodies should signal quickly how assessment and accountability frameworks will evolve so teachers can plan with confidence rather than uncertainty.

If the UK matches the technical and regulatory ambition expressed elsewhere — and pairs it with sustained investment in teachers and pupils — the outcome could be broadly positive. Across the surveys and reports, a consistent theme emerges: people want the next generation to understand AI and to be prepared for an AI-rich future. They are not opposed to innovation; they want it done responsibly. That is a clear opportunity for education leaders and policymakers to shape a model of AI in schools that enhances learning, preserves human judgement where it matters most, and sets the UK up as a global exemplar in responsible edtech.

Source: Noah Wire Services

More on this

https://opinium.substack.com/p/you-cant-brush-ai-under-the-carpet - Please view link - unable to able to access data
https://www.opinium.com/public-attitudes-to-the-ai-opportunities-action-plan/ - Opinium’s analysis of public attitudes to the UK government’s AI Opportunities Action Plan finds mixed support for AI across sectors, with acceptance depending on task risk. The public favour AI handling repetitive administrative duties but express scepticism about AI planning lessons or making high-stakes educational decisions. Younger people show greater openness to AI than older cohorts. Key public conditions for trust include proven accuracy, regulation, accountability and data security. The report highlights that perceptions vary by task — administrative automation attracts more approval than assessment — and recommends transparency and safeguards to build confidence in AI’s educational role nationwide adoption.
https://schoolsweek.co.uk/public-divided-over-ai-use-in-schools/ - Schools Week reports on a recent poll revealing the public is divided over AI use in schools. The piece notes majority support for teachers using AI for administrative tasks but strong opposition to AI marking student work. It details gender differences in attitudes and highlights calls for national strategy and funding to ensure equitable AI adoption. The article quotes education leaders urging adaptation of coursework to remain meaningful in an AI era and reports government guidance recommending schools plan for AI use while considering safeguarding and privacy. The coverage frames public sentiment as cautious yet open to well-governed AI deployment.
https://www.gov.uk/government/publications/parent-pupil-and-learner-voice-omnibus-surveys-for-2024-to-2025/parent-pupil-and-learner-voice-december-2024 - GOV.UK’s omnibus survey reports on pupils’ use of generative AI and parental attitudes, showing many secondary pupils have used AI tools in lessons and for homework. The dataset breaks down usage by key stage, showing increases since prior waves. It presents figures for in-school and homework use and highlights differences between cohorts. The publication emphasises policymakers should account for widespread familiarity with AI among young people when designing guidance and curriculum. The report underlines the need for schools to develop clear policies on safe, ethical and pedagogically sound AI use within teaching and assessment frameworks for learners and staff alike.
https://thejournal.com/articles/2025/05/14/survey-ai-optimism-is-rising-but-cheating-and-privacy-concerns-persist.aspx - THE Journal’s survey of US educators and administrators finds rising optimism about AI’s potential in K‑12, coupled with persistent concerns about cheating, privacy and academic integrity. Most respondents report increased AI use for drafting communications and lesson materials, and many see benefits in reducing administrative workload. Yet the survey highlights that cheating with AI is a major worry, with a large proportion experiencing incidents and noting inadequate detection tools. The article stresses the need for clearer policies, better training and robust privacy safeguards. It concludes that while AI adoption is accelerating, confidence depends on effective governance and educator support systems.
https://www.brownejacobson.com/about/news-media/school-leaders-survey-illustrates-how-teachers-are-adopting-ai - Browne Jacobson’s report on its School Leaders Survey outlines how many school leaders are adopting AI tools to ease workload, with around half reporting some use and one in five using tools regularly. Respondents highlighted improvements in efficiency, lesson preparation and personalised learning, but also flagged gaps in AI expertise across organisations. Concerns include plagiarism, safeguarding and legal compliance. The piece notes government guidance urging safe deployment and recommends training and planning to ensure equitable access. The survey underlines that while AI can assist administrative and planning tasks, schools need clear policies and capacity-building to manage risks and maximise benefits.
https://www.cambridge.org/news-and-insights/public-divided-on-ai-in-schools - Cambridge University Press & Assessment summarises a nationally representative survey showing the public supports teachers using AI for administrative tasks but largely opposes its use for marking. The piece reports a survey of over 2,200 UK adults and notes nuanced attitudes: many back AI for routine tasks yet distrust it with assessment responsibilities. It highlights age-related differences, with younger respondents more receptive to AI. The article discusses implications for policy and calls for evidence, regulation and teacher training. It frames the debate as balancing AI’s potential to ease burdens against preserving human judgement in core educational functions and professional accountability.

Noah Fact Check Pro

The draft above was created using the information available at the time the story first emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed below. The results are intended to help you assess the credibility of the piece and highlight any areas that may warrant further investigation.

Freshness check

Score: 8

Notes: The narrative presents recent polling data from Opinium, conducted between 6 and 8 August 2025, indicating a high freshness score. However, similar discussions on AI in education have been covered in The Observer, such as the article 'Question: who is marking your GCSE? Answer: an AI examiner' published on 15 August 2025. ([observer.co.uk](https://observer.co.uk/news/national/article/question-who-is-marking-your-gcse-answer-an-ai-examiner?utm_source=openai)) This overlap suggests that while the polling data is fresh, the broader topic has been recently discussed. Additionally, the narrative appears to be based on a press release, which typically warrants a high freshness score. Nonetheless, the presence of similar content within a short timeframe raises questions about originality.

Quotes check

Score: 9

Notes: The narrative includes direct quotes from Opinium's polling data. A search for the earliest known usage of these quotes indicates that they originate from the recent Opinium survey conducted between 6 and 8 August 2025. No identical quotes appear in earlier material, suggesting that the quotes are original to this survey. This originality enhances the credibility of the content.

Source reliability

Score: 7

Notes: The narrative is based on a press release from Opinium, a market research firm. While Opinium is a known entity, its press releases are self-published and may present information in a manner favourable to their interests. The Observer, a reputable UK newspaper, has covered similar topics, such as AI's role in education, indicating that the subject matter is of public interest. However, the reliance on a self-published press release introduces potential biases, warranting a moderate reliability score.

Plausability check

Score: 8

Notes: The claims regarding public sentiment on AI in education align with recent discussions in reputable outlets like The Observer. The narrative's tone and language are consistent with UK English usage, and the structure is coherent without excessive or off-topic details. However, the overlap with similar content within a short timeframe raises questions about the originality of the narrative.

Overall assessment

Verdict (FAIL, OPEN, PASS): OPEN

Confidence (LOW, MEDIUM, HIGH): MEDIUM

Summary: The narrative presents fresh polling data from Opinium, conducted between 6 and 8 August 2025, indicating a high freshness score. The quotes are original to this survey, enhancing credibility. However, the reliance on a self-published press release introduces potential biases, and the overlap with similar content in reputable outlets like The Observer raises questions about originality. Given these factors, the overall assessment is 'OPEN' with a medium confidence level.

Education
Artificial Intelligence
Policy