A school decides whether to pilot an AI tutoring tool
A secondary school's leadership must decide whether to run a one-term pilot of a commercial AI tutoring tool for pupils aged 11 to 16.
Setting. A publicly funded secondary school with a head teacher, a governing board and one part-time IT coordinator. The vendor offers a discounted licence for the pilot. Teachers are stretched, and the school hopes the tool will give pupils help outside lessons. No real school, vendor or jurisdiction is assumed.
The outcome at stake is a pattern, not one pupil's homework: whether pupils learn more, or learn to hand the thinking over. A tool that answers well can still reduce the effort that builds skill. The benefit is real too: help for pupils whose families cannot pay for tutoring.
Unknown in this scenario
Whether the pilot will measure learning, or only use and satisfaction.
source-assertionConfidence: mediumContestability: high
In a survey of knowledge workers published in 2025, people who trusted generative AI more reported doing less critical thinking, and their effort shifted toward checking and integrating AI output.
Our reading: Self-reported and correlational, with no measure of performance; most authors work for a company that sells these tools. A study of office workers does not describe pupils or teachers.
Inspect evidence · 1 passage
Passage summaries are paraphrases, not quotations. Open the source to read the full passage.
In a survey of 319 knowledge workers describing 936 uses of generative AI, higher confidence in the AI was associated with less self-reported critical thinking, and higher self-confidence with more. Critical thinking shifted toward verifying and integrating AI output.
Scope note: Six of seven authors work at Microsoft, which sells generative AI tools. Self-reported, cross-sectional survey; it measures perceptions, not performance. The ACM page was not reachable; the passage was read on the Microsoft Research publication page.
Requirement 2, “Ensure safety for children” (about p. 15)
Children’s safety in relation to AI systems must be assured in both the short and the long term.
Scope note: Supersedes the 2021 Policy guidance on AI for children (2.0). Ten requirements for child-centred AI; guidance, not law. Page numbers below are from the PDF as extracted.
What incentives currently produce or reinforce that outcome?
The vendor is rewarded for adoption and for time spent in the tool. The school is rewarded for results and for relieving teachers. A discounted pilot makes the next year's decision harder to refuse. Pupils are rewarded for finished homework, not for the effort behind it.
Unknown in this scenario
The price after the pilot.
Whether the vendor's contract rewards it for learning outcomes or only for use.
The head teacher can approve a pilot; the governing board approves spending above a threshold; the vendor controls the model and its updates; the education authority sets data-protection rules. No one at the school has authority over how the model is trained.
Unknown in this scenario
Whether parents have any formal route to object.
Whether the education authority must approve tools that process pupils' data.
Does that institution have the capacity, independence, expertise, legitimacy, and accountability to act?
The school can watch how pupils use the tool, but it cannot test the model independently. Its only technical capacity is a part-time coordinator. Most of what it knows about accuracy and safety comes from the vendor, so its independence is limited.
Unknown in this scenario
Whether a sector body evaluates or lists approved tools the school could rely on.
source-assertionConfidence: highContestability: medium
UNESCO’s 2023 guidance recommends thirteen as the minimum age for independent conversations with generative AI and strict validation of tools before schools or universities officially adopt them.
Our reading: The guidance itself notes that some consider thirteen too young. Validation requires capacity many institutions lack.
Inspect evidence · 1 passage
Passage summaries are paraphrases, not quotations. Open the source to read the full passage.
The guidance recommends thirteen as the minimum age for independent conversations with generative AI, noting that some consider this too young, and calls for strict ethical and pedagogical validation of tools before institutions officially adopt them.
Scope note: The UNESCO repository page did not render; the passage was read in an open copy held by University College London.
What infrastructure does the mechanism depend on or reshape, and who controls it?
The tool depends on pupils' devices and home internet, the vendor's cloud service, and the data pupils type into it. The vendor controls all but the devices. Pupils without reliable access at home get less of the benefit.
Unknown in this scenario
Where pupils' data is stored, for how long, and whether it is used to train the vendor's models.
How many pupils lack a suitable device or connection at home.
What safeguards translate the objective into practice?
Possible safeguards include default settings that give hints rather than answers, a clear route for any pupil to reach a teacher, and notice to parents. Others are a data agreement and an end-of-term review with stop criteria decided in advance. Each exists only once someone is named to carry it out.
Unknown in this scenario
Who reviews concerns that pupils or parents raise during the pilot.
Where in the AI lifecycle does the intervention operate?
The decision sits at procurement (buying) and deployment (putting it into use). The pilot also needs monitoring during the term. It needs a plan for switching off: what happens to pupils’ data and to lessons built around the tool if the school stops.
Unknown in this scenario
Whether the contract requires the vendor to delete pupils' data at the end of the pilot.
Who decided, on what authority, and who was not represented?
Leadership and the vendor decide. Pupils, who will live with the result, cannot meaningfully consent and are not asked by default. Parents, the teachers who will supervise, and pupils with disabilities or who speak another language at home are absent unless the school invites them.
Unknown in this scenario
Whether pupils or parents were consulted before the pilot.
UNESCO’s guidance recommends 13 as the minimum age for independent use of generative AI; whether the pilot will include pupils aged 11 and 12, and on what terms.
source-assertionConfidence: highContestability: medium
UNESCO’s 2023 guidance recommends thirteen as the minimum age for independent conversations with generative AI and strict validation of tools before schools or universities officially adopt them.
Our reading: The guidance itself notes that some consider thirteen too young. Validation requires capacity many institutions lack.
Inspect evidence · 1 passage
Passage summaries are paraphrases, not quotations. Open the source to read the full passage.
The guidance recommends thirteen as the minimum age for independent conversations with generative AI, noting that some consider this too young, and calls for strict ethical and pedagogical validation of tools before institutions officially adopt them.
Scope note: The UNESCO repository page did not render; the passage was read in an open copy held by University College London.
Who bears the costs and risks, and who captures the benefits?
Teachers absorb extra supervision, pupils supply their data, and pupils least able to spot a wrong answer carry most of the risk of one. The vendor gains a reference customer and usage data. Pupils with home access gain the most help.
Unknown in this scenario
Whether benefits reach the pupils who most need help, or mainly those already doing well.
What happens when the mechanism or technology scales across organizations, sectors, or borders?
If many schools adopt the same tool, one vendor shapes how a generation practises learning, and its errors repeat everywhere at once. Sector bodies may end up as the real gatekeepers, because single schools cannot evaluate tools.
Unknown in this scenario
Whether any sector body would see problems across schools early enough to act.
What becomes difficult to reverse if governance arrives too late, and how will the governance mechanism itself be revised if it turns out to be wrong?
After a year, homework, lesson plans and pupils' habits may be built around the tool, and the vendor holds a year of data. Keeping the choice reversible means deciding now how the pilot will be reviewed, who can stop it, and how the school would leave.
Unknown in this scenario
Whether the school would still have the staff time to teach without the tool after a year.
The technical question asks whether the tutor gives good answers. The governance questions show a school that controls defaults and access but not the model. The pilot is priced so that continuing is easier than stopping. It affects children who were not asked, and its safeguards exist only if someone is given the job. Most important facts are unknown at the point of decision, so the pilot should be designed to find them out.
CHI 2025 (ACM) · scholarship · published 2025-04-25
Six of seven authors work at Microsoft, which sells generative AI tools. Self-reported, cross-sectional survey; it measures perceptions, not performance. The ACM page was not reachable; the passage was read on the Microsoft Research publication page.
lee-2025-genai-critical-thinking · Last checked 2026-10-18 · claims and passages
UNICEF Innocenti – Global Office of Research and Foresight · guidance · publication date not recorded · 3.0; published December 2025
Supersedes the 2021 Policy guidance on AI for children (2.0). Ten requirements for child-centred AI; guidance, not law. Page numbers below are from the PDF as extracted.