Generative AI And Closing The Critical Literacy Skills Gap

The role of generative AI in education

Interest in and use of artificial intelligence (AI) has exploded in the past year.

The rapid development of generative AI will undoubtedly continue to gain momentum. The evolution of the digital sphere and the rise of new, readily available technologies have made us consider some of the questions that naturally arise around the relationship between AI and literacy.

This evolving literacy and education landscape is nothing new. From the invention of writing to the invention of generative AI, what it means to be literate has always been changed by the technology we use to communicate. Today, being literate increasingly includes the skills of effectively, critically and creatively engaging with generative AI, yet our education system is struggling to keep pace with its rapid growth and capabilities.

We know that, universally, there are mixed feelings about AI as we all grapple with the ever-changing technology and begin to better understand how to harness its power. This is perhaps most pertinent to teachers and educators who are trying to get to grips with the technology their students are more comfortable using.

There are also many questions to consider about how future generations may harness these digital tools to support their learning and what this means for their future literacy.

What do we know about generative AI and literacy so far?

The potential impact of generative AI, including tools such as ChatGPT, Claude, Gemini, Midjourney and Sora, is of particular interest to those working in the creative industries and education.

In terms of education, experts suggest that AI could help to address big educational challenges – such as reducing teacher workloads and providing new opportunities for those with additional needs – but that it will require supporting people to learn about its safe and effective use and may mean moving the focus of education onto human intelligence to ‘prepare people for an AI world’ (Professor Luckin, 2019).

The ability of generative AI tools to both write and read texts in a human-like manner means they are set to play an increasingly important role in the literacy lives of children, young people and adults. Indeed, academics have suggested that ‘literacy increasingly means and includes interacting with and critically evaluating AI’ (McKnight, 2021).

To this end, we were interested in exploring how generative AI tools might influence, and potentially redefine, what it means to be literate in the digital age. As a first step, we surveyed tens of thousands of children and young people – as well as 1,200 primary and secondary teachers – in 2023 and 2024 to explore attitudes, behaviours and confidence around using generative AI to support literacy and learning.

What children and young people told us

Our reports show a dramatic increase in the percentage of young people aged 13 to 18 who say they have used generative AI over the last year. In 2024, more than 3 in 4 (77%) said they had used it, doubling from 2 in 5 (37.1%) in 2023.

How are children and young people using generative AI?

Most young people who used generative AI regularly told us they were using it for entertainment, curiosity, homework and inspiration. However, many were also using it for literacy-related purposes.

More than 2 in 5 (44%) told us they were using it to have a chat, 1 in 5 (19%) to write stories and 1 in 8 (13%) to write poems or lyrics. More generally, 2 in 5 (40%) said it helped them with writing and almost 1 in 4 (23%) with reading.

Interestingly, with our recent research also revealing a crisis in children and young people’s writing engagement, generative AI was found to provide potential routes into writing for some who don’t enjoy it, with 1 in 7 (14.5%) reluctant writers saying they use AI to write stories.  

What are children and young people’s attitudes towards generative AI?

Given wider concerns about increased plagiarism and potential bias and inaccuracy relating to generative AI use, we also asked young people about their attitudes and behaviours when using these tools.

Almost half (47%) of the young people in our survey told us that they usually added their own thoughts into anything generative AI told them and 2 in 5 (40%) said they checked AI outputs as they could be wrong. However, 1 in 5 admitted that they usually just copied what it told them (21%) and that they didn’t check outputs (21%).

This suggests that many young people would benefit from greater support to work effectively with generative AI and to learn the skills they need to critically evaluate AI outputs. This is further supported by more than 4 in 5 (82%) teachers agreeing that young people should be taught how to engage critically with generative AI tools.

What teachers told us

As with young people, we saw a rise in the percentage of teachers saying they had used generative AI tools over the last year. In 2024, 1 in 2 (48%) teachers said they’d used these tools (31% in 2023).

Interestingly, while more secondary than primary teachers use generative AI, the percentage of secondary teachers using it has declined in the past year (from 76% in 2023 to 57% in 2024). By contrast, its use has increased with primary teachers (from 20% in 2023 to 31% in 2024).  

Most teachers who used generative AI regularly were using it for similar reasons to young people (curiosity, ideas and questions) but many also used it to support their work. For example, 2 in 5 (38%) teachers had used it to create lesson content and 1 in 3 to generate model answers (35%) or for lesson planning (32%).

At the same time, nearly 2 in 5 (38%) teachers were concerned about their pupils using generative AI. Indeed, most of the teachers’ responses to our survey reflected the mixed feelings many of us have about the potential benefits and drawbacks of these new technologies. For example, while 2 in 3 (65%) teachers agreed that generative AI could model good writing, 1 in 2 (49%) also believed it could have a negative impact on children’s writing skills overall.

Similarly, while nearly 3 in 5 (57%) teachers were concerned that generative AI could stop children thinking for themselves, and 2 in 5 (42%) felt it could decrease children’s engagement with learning, more than half (56%) felt that students who weren’t supported to use generative AI effectively would be at a disadvantage in the future workplace.

Notably, it’s not just young people who teachers feel need more support to engage critically with generative AI tools – 3 in 4 (75%) wanted more training to use it effectively themselves.

What are the pros and cons of using AI in the classroom?

There are clearly many applications for generative AI both in and outside of the classroom. However, it is perhaps the debate around the pros and cons of its usage which needs to be addressed. As with all new technologies and developments, there is more information to be gleaned over time which will support more informed decisions. But let us start to explore some of the pros and cons based on our initial research.

It is worth noting that as with most balanced arguments, there can be shades of grey within opinions and space for a sliding scale where positive and negative outcomes can still be held as possibilities.

Con: Young people need better critical literacy skills to use generative AI effectively

Our findings show that while many young people who use generative AI say they add their own thoughts and check outputs, a fifth admit they do not interact with these tools in a critical and creative way. The majority of teachers also raised concerns that young people do not have the critical literacy skills needed to effectively and safely use generative AI tools.

This echoes earlier research we conducted for the Parliamentary Commission on Fake News and the Teaching of Critical Literacy Skills, where only 2% of children and young people were found to have the critical literacy skills needed to correctly identify real and fake news and where half of teachers (54%) believed the national curriculum did not equip children with the critical literacy skills they need to identify fake news.

Online environments are an integral part of the lives of many young people. They can offer valuable opportunities for literacy, communication, self-expression and support for many young people. However, while online communication platforms often require young people to employ foundational literacy skills (including reading, writing, speaking and active listening), broader and more complex literacy skills (including critical digital literacy) are equally important.

It is essential that all young people build the critical digital literacy skills they need to learn, not just in terms of how to use the tools themselves but also about their wider digital ecosystem. This includes, for example, the ethical and environmental implications of generative AI, inaccurate content and potential biases inherent in training data.

In addition, our previous research found that literacy engagement was associated with positive critical digital literacy attitudes and behaviours, which in turn were linked with greater mental wellbeing. 

Con: Generative AI could cause poorer writing skills overall

While many teachers can see the potential benefits of generative AI for modelling good writing, they also feel that coming to depend on it could cause poorer writing skills overall.

Similarly, while we may be pragmatic about the need for young people to learn about the effective use of these tools for the future workplace, many teachers are concerned about decreased engagement with learning (such as when online research is analysed and synthesised on their behalf).

Con: Teachers are unsure of how to use generative AI effectively and many aren’t being supported to do so

Most teachers feel that while young people need support, they too would benefit from training and resources to help them understand how generative AI might be used most effectively to support literacy in the classroom. This extends to their whole-school approach, with only 1 in 10 (10%) teachers saying their school has an AI policy in place.

While many are starting to recognise some of the potential advantages of generative AI for their roles, few are putting the tools into practice. For example, half (50%) of teachers agree generative AI could support assessment but very few (5%) are using it for this purpose.

Pro: Strong literacy skills support more effective use of generative AI

As generative AI tools become more commonplace, some commentators have suggested that foundational literacy skills (reading and writing) will be vital for getting the most out of what they can offer.

Reading and writing skills enable and support users to communicate effectively with generative AI (i.e. to craft a good question or prompt) and to critically assess the quality, reliability and ethics of the generated content. Learning to work effectively with AI tools and responses (e.g. adding your own take on the content generated, checking sources and deciding what to keep) is part of developing good AI literacy skills, which have been defined as ‘understanding, using, evaluating and [considering] ethical issues’ relating to AI (Ng et al., 2021).

What are the best AI tools for teachers?

Whether you lack the confidence to use AI in the classroom or are chomping at the bit to get started, there is a wealth of applications and resources already available if you are not certain what AI tools are best for you.

We believe that it is crucial to understand which tools can be accessed so that, as educators, you feel confident to equip all children and young people with the critical and digital literacy skills they need to thrive in education, work and society – both today and in the future.

Several organisations have developed resources and tips to help teachers use generative AI to support literacy in and out of the classroom. Before using these, we recommend reading the guidance about minimising harm in the UK Government’s generative artificial intelligence in education document.

(Please note: these examples are for illustrative and informational purposes only; their inclusion does not imply endorsement.)

Useful tips, resources and AI tools

• Check your school’s AI policy, if there is one. If you don’t have an AI policy, it’s a great idea to get governors, senior management teams, parents and young people together to develop one. Here is a template by Mark Anderson and Laura Knight to help you get started.
• Check if your school already uses built-in generative AI tools. These might include Microsoft Copilot for Bing or Google’s Gemini. These tools are a good place to try out generative AI if you’re a beginner. You should have a look at what other support is offered too. Some schools may have signed up to specific platforms such as TeacherMatic, which also offers limited free options for access.
• Remember: never submit confidential, sensitive or personal information to any AI platform.
• Be aware that generative AI can make mistakes – always remember to check outputs.
• The PAIR framework is a good approach for trying out new generative AI tools.

AI tools to support primary and secondary teaching:

• OpenAI has provided some tips for supporting teaching with generative AI.
• Oak Academy is piloting generative AI lesson planners and quizzes suitable for primary and secondary teachers.
• Mark Anderson (@ICTevangelist) has produced a free and accessible PDF of ChatGPT prompts for teachers of various phases and subjects.
• The Education Endowment Foundation (EEF) is currently conducting research designed to measure the potential for generative AI to reduce teacher workload, with an emphasis on secondary STEM teaching.

Examples of the ways generative AI tools can support your teaching:

• Explore story generation, character development or role-playing/dialogue with book characters.
• Craft narratives or non-fiction writing based on children’s interests and abilities.
• Support summarising or differentiating texts.
• Develop comprehension quizzes or cloze procedures.

---

Irene Picton is Senior Research Manager at the National Literacy Trust.