In my previous reflection on curriculum design, I argued that we should not begin by asking where AI can be inserted. We should begin with the graduate: what competence students must demonstrate, what thinking must remain theirs and how assessment will make that learning visible.
That creates another question. If lecturers decide where AI belongs in a curriculum, how it may be used in assessment and what responsible student use looks like, how AI literate do lecturers need to be? My answer is: sufficiently literate to make sound educational and professional decisions. They do not need to be programmers, data scientists or IT experts.
We cannot design responsibly around something we do not understand
Students are already using generative AI. Lecturers are being asked to write policies, redesign assessment, approve learning activities, evaluate AI-generated materials and advise students about acceptable use.
Those decisions require enough understanding to judge what a system can do, where it may fail, whether it supports or bypasses an intended outcome and what students must still demonstrate independently.
AI literacy is not technical expertise
Lecturers do not need to build machine-learning models, understand advanced coding or become familiar with every AI product. They do need a working understanding of how generative AI produces responses, why it can fabricate information, how bias can enter the output and why fluent language is not evidence of accuracy.
The goal is not technical mastery. It is informed judgement.
The literacy lecturers actually need
- Foundational: capabilities, limitations, hallucinations, bias and verification.
- Pedagogical: connecting AI use to intended outcomes and protecting the cognitive work students must perform.
- Assessment: designing AI-free, AI-supported and AI-critical tasks that make reasoning visible.
- Ethical and professional: privacy, confidentiality, copyright, fairness, accessibility, transparency and accountability.
- Discipline-specific: recognising relevant uses and risks within physiotherapy practice and education.
- Reflective: continuing to test approved tools, examine evidence and learn with colleagues.
Not every lecturer needs the same level
An AI-aware lecturer can explain basic concepts and risks, communicate policy and guide responsible student use. An AI-enabled lecturer can design aligned activities, adapt assessment and evaluate generated resources. An AI curriculum leader can coordinate programme design, mentor colleagues and contribute to institutional strategy.
The aim is not to make every lecturer a curriculum leader. It is to establish a safe baseline for everyone and develop deeper capability where particular roles require it.
Universities must develop people before platforms
Institutions sometimes move quickly to purchase tools while faculty development remains optional or is squeezed into already crowded workloads. An AI-ready university needs protected development time, approved platforms, clear policy, practical examples, assessment support and communities of practice.
If lecturers remain responsible for curriculum quality and professional standards, they need genuine opportunities to develop the literacy those responsibilities require.
The missing step in curriculum design
My earlier sequence was: CILO → Human competence required → Permitted AI role → Learning activity → Evidence of student reasoning → Assessment and assurance.
I would now add one step: Lecturer competence required. An AI-ready curriculum requires AI-ready educators. This does not mean becoming computer scientists. It means making defensible decisions about when, why and how AI should—or should not—be part of learning.
How AI literate are you?
I created a short 15-question knowledge check for students, lecturers and other health-professions learners. It covers verification, privacy, bias, professional responsibility and appropriate educational use.
The quiz is a reflective learning activity, not a validated assessment. Results are stored only in your browser and displayed as local averages by participant role.