Muhammad Tayyab Ilyas

Chapter 8. Safe, Fair and Honest

During a discussion about online privacy, a ten-year-old girl explained how she protected herself. She used a name she had invented. ‘I make up a name’, she told the researchers (Zhao et al., 2019, p. 7).

Her answer showed care. She understood that people online did not need to know everything about her. Other children described checking someone’s identity in person before accepting an online request. Yet the group understood much less about information that apps collected quietly. Familiar apps could feel safe because nothing bad had happened before.

These discussions took place with primary children in the UK, before generative AI became widely available. They offer emerging evidence about children’s thinking, from a small study. We cannot infer how those children would use a chatbot today. The distinction still matters. A child can follow the rule about names while adults overlook the system collecting information around her.

What the chapter covers

  • Data, permission and monitoring
  • Fair judgement and honest explanations
  • Safeguarding beyond the lesson
  • In practice
  • When resources are thin
  • Questions for reflection
  • Framework connections

Questions for reflection

Bring these questions to a phase or department discussion, keeping real students’ details out of the conversation.

  1. Think of an AI activity you plan for your own class: what information about your students does it use, and could they learn the same thing without it?
  2. Who receives a safeguarding alert in your school, and what happens when that person is unavailable?
  3. What independent evidence would let you reject an AI recommendation about a student?
  4. How could a student question a judgement or disclose permitted assistance without feeling accused?
  5. What equivalent learning route can a family choose if they decline direct use of an AI service?