Explain what this recent report supports, how its evidence was gathered, and what a careful reader should not infer from it.
A new pattern raises a useful question
When students say they use an AI tool, what can that tell us about learning? The National Literacy Trust’s August 18, 2026 report offers evidence about use and reported attitudes, but those are not the same outcome as learning. The report draws on early-2026 surveys of young people, teachers, and parents, along with focus groups and interviews. A careful essay can use this release to discuss how AI has entered literacy practices while keeping a boundary between what respondents reported and what a separate learning assessment would need to measure.
Name who answered
The report’s sample includes 40,543 young people aged 13 to 18, 2,567 teachers, and 3,073 parents, plus qualitative conversations with young people and adults. The headline that 79.0% of surveyed young people had used generative AI refers to the surveyed group and the survey question; it is not a direct count of every young person in every country. Identifying respondents early makes the claim more accurate. It also lets a reader ask how the survey recruited participants, how questions were phrased, and whether the result should be generalized beyond its intended population.
In the margins Why this passage works +
“The report’s sample includes 40,543 young people aged 13 to 18, 2,567 teachers, and 3,073 parents”
This passage identifies the sampled groups and keeps the scope visible before the essay interprets a headline percentage.
Separate frequency from purpose
The report says 59.5% of surveyed young people used generative AI weekly or more often. It also describes reported purposes, including understanding new topics, getting started on work, and receiving initial writing feedback. These findings answer different questions: frequency describes how often, while purpose describes what respondents say they are doing. A sound paragraph should not combine them into the vague claim that “students rely on AI for school.” More precise wording could say, “In the National Literacy Trust survey, many respondents reported using generative AI regularly, with some describing it as a way to begin or get feedback on writing.”
Describe change without inventing a cause
The report also says the share of young people who reported checking AI outputs increased from 42.8% in 2025 to 52.3% in 2026. This comparison supports a statement about a change in self-reported checking within the survey series. It does not tell us why the change occurred. Perhaps participants developed greater experience, teachers emphasized verification, or survey conditions differed; those are possible explanations, not findings established by the percentage alone. An essay can introduce a question about causes, but it should label the question as a question and look for additional evidence before answering it.
In the margins Why this passage works +
“It does not tell us why the change occurred.”
The explicit boundary prevents a before-and-after difference in reported behavior from being turned into a causal explanation.
Use mixed methods with care
Survey percentages can show how common a response was among respondents, while focus groups and interviews can help explain how participants describe their choices. Neither method substitutes for the other. A selected comment illustrates a possible experience; it does not show how common that experience is. A percentage summarizes a response pattern; it does not capture every reason behind it. The report’s combination of methods gives writers more than one kind of evidence to discuss, but a careful analysis still identifies which kind supports each sentence. This distinction is useful well beyond AI: it applies whenever a report mixes counts with participants’ words.
What the report does not establish
Self-reported use is not a test score. From this release alone, a writer cannot conclude that AI improved writing, weakened independent thinking, or caused a change in literacy. Those claims would require measures of learning or writing quality and a design capable of comparing outcomes under relevant conditions. Instead, an evidence-aligned conclusion might be that AI use was common among the surveyed young people and that some respondents described increasingly reflective practices, including checking outputs. That conclusion is less sweeping than a headline about transformation, but readers can see exactly how the report supports it.
A sentence frame for critical reading
Try a two-part summary: “In a 2026 survey and qualitative study, the National Literacy Trust found that [respondents] reported [specific pattern]. Because the report measured [self-reported behavior or perception], it does not by itself establish [learning or causal claim].” Fill the brackets with the population and measure actually described in the source. Then ask what additional research would answer the question you still have. The method teaches a general habit: name the measure before interpreting the result, and place the limitation next to the claim it qualifies.
Try the idea.
Write a two-sentence summary of the report. Sentence one should identify who was surveyed and one behavior they reported. Sentence two should name one conclusion the survey alone cannot establish. Underline the phrase that signals the limit.
See the self-check +
A strong response identifies the respondent group and a specific self-reported measure, then distinguishes that measure from a causal or learning outcome not tested by the survey.
Sources and evidence
- Young people, teachers’ and parents’ use of AI to support literacy in 2026 · National Literacy Trust
- Annual Literacy Survey · National Literacy Trust