Every week another headline asks whether artificial intelligence (AI) will replace teachers.
I think we are asking the wrong question.
The real question is this: Are we preparing future teachers to know when AI is right and when it is wrong?
AI is becoming part of education.
Teachers are using it to generate lesson plans, write assessments, develop rubrics and to create classroom resources in minutes.
The temptation is to believe that because AI is fast, it must also be accurate.
It isn’t.
My recent research explored how preservice teachers used AI to design curriculum aligned assessments.
What I found was encouraging but also concerning.
Most student teachers embraced AI because it saved time and generated creative ideas.
They used tools such as ChatGPT, Conker AI and other platforms to develop questions, assessment structures and learning activities.
AI reduced the burden of starting from a blank page and helped spark new ideas.
But that was only part of the story.
The strongest student teachers did something AI can never do on its own.
They questioned the answers.
They recognised when AI generated content at the wrong grade level.
They noticed when activities were not aligned with the SA curriculum.
They rewrote questions, simplified language, considered pupils’ needs and reflected on whether the assessment was fair and appropriate.
They understood that technology should support teaching not replacing professional thinking.
This matters because AI is not neutral.
It has no understanding of the pupils sitting in front of a teacher.
It does not know the realities of overcrowded classrooms, limited resources or diverse learning needs.
It cannot judge whether a question is culturally appropriate or whether an assessment truly measures understanding.
Teachers can.
That is why I argue that teacher education needs to move beyond AI literacy.
Knowing how to write a good prompt is no longer enough.
Future teachers must learn how to evaluate AI generated content critically, identify inaccuracies, recognise bias, protect pupil privacy and make professional decisions about what should be accepted, adapted or rejected.
In my research, I describe this as AI pedagogical judgement.
It is the professional ability to balance efficiency with ethics, curriculum requirements with creativity and technological possibilities with sound educational practice.
This shift has important implications for higher education.
If we continue teaching AI as merely another digital tool, we risk producing graduates who rely on technology rather than their own professional expertise.
Instead, teacher education programmes should provide structured opportunities for students to critique AI outputs, justify their decisions and reflect on how technology influences teaching and assessment.
The future of education will include AI.
But classrooms still need teachers who can think critically, exercise sound judgement and make decisions in the best interests of pupils.
The future does not belong to teachers who simply know how to use AI.
It belongs to teachers who know when not to trust it.
Warda Abrahams, senior lecturer- teaching and learning, Emeris Nelson Mandela Bay
Follow The Herald WhatsApp channel today and stay connected to the stories shaping our world.
The Herald
