{"id":321,"date":"2026-09-12T07:30:00","date_gmt":"2026-09-11T23:30:00","guid":{"rendered":"https:\/\/oribridge.org\/blog\/?p=321"},"modified":"2026-09-08T09:37:14","modified_gmt":"2026-09-08T01:37:14","slug":"ai-teacher-training-role-map-not-generic-curriculum","status":"publish","type":"post","link":"https:\/\/oribridge.org\/blog\/ai-teacher-training-role-map-not-generic-curriculum\/","title":{"rendered":"AI Teacher Training Has a Role Map, Not One Generic Curriculum"},"content":{"rendered":"

An overseas education expert is asked to adapt an AI course for China. The first draft has one audience label: \u201cteachers and education leaders.\u201d It has one sequence of lessons, one set of exercises and one outcome statement about using AI responsibly.<\/p>\n

That draft may be well intentioned and still be too broad. Chinese government teacher-development documents describe several roles and capability-building contexts. The Ministry of Education\u2019s digital empowerment action discusses teacher digital literacy, national training channels and AI-focused training for teachers and school leaders. The \u201cAI+Education\u201d action plan also refers to training, standards, evaluation and differentiated capability building.<\/p>\n

These documents are policy signals, not a purchase order for one foreign course. Their practical value is that they warn an expert to map the role before adapting the curriculum.<\/p>\n

The word \u201cteacher\u201d hides several decisions<\/h2>\n

An education administrator may need to decide how an institution sets priorities, supports staff and evaluates a programme. A school leader may need to organise a local implementation plan. A classroom teacher may need to design or review an activity. A researcher or trainer may need to compare practices and develop guidance.<\/p>\n

Those roles can discuss the same technology and still require different work. A generic lecture about prompt writing may be too operational for a policy audience and too abstract for a classroom teacher. A lesson about school-wide governance may be useful to a leader but irrelevant to a teacher who needs to create and review a specific learning activity.<\/p>\n

The policy pages do not tell an overseas expert which role is present in a particular buyer conversation. They do justify asking the question before translating a curriculum.<\/p>\n

Read policy as a role map, not a syllabus<\/h2>\n

The Ministry of Education\u2019s teacher-development action places digital literacy and AI-related training inside a broader teacher-development system. The AI+Education action plan describes education, training, application innovation, standards and support as connected areas, and mentions differentiated capability development.<\/p>\n

An expert can use those documents to create a scoping table:<\/p>\n

\n\n\n\n\n\n\n\n\n
Role<\/th>\nDecision or work task<\/th>\nPossible evidence of learning<\/th>\n<\/tr>\n<\/thead>\n
Education administrator<\/td>\nChoose a capability priority and define institutional constraints<\/td>\nA short priority memo with stated assumptions<\/td>\n<\/tr>\n
School leader<\/td>\nDesign a bounded implementation or support plan<\/td>\nA role\/responsibility map and review checkpoint<\/td>\n<\/tr>\n
Teacher<\/td>\nBuild or review one AI-supported learning activity<\/td>\nAn annotated activity and limitation note<\/td>\n<\/tr>\n
Trainer\/researcher<\/td>\nCompare practices and formulate guidance<\/td>\nA source-backed comparison with uncertainty marked<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

The table is an internal design aid, not a government taxonomy or a claim that all buyers use these labels. Its purpose is to stop the curriculum from treating every participant as the same learner.<\/p>\n

A concrete scene: one workshop, four wrong outcomes<\/h2>\n

Imagine a two-hour workshop titled \u201cUse Generative AI in Education\u201d. The organiser invites a principal, a curriculum researcher and classroom teachers. The instructor demonstrates a chatbot, asks everyone to write a prompt and ends with a general discussion about ethics.<\/p>\n

At the end, the principal wants a school plan, the researcher wants a comparison framework, and the teachers want a classroom activity they can inspect. Nobody is necessarily dissatisfied with the demonstration. The problem is that the workshop never chose a work product.<\/p>\n

A role map would expose the collision before delivery. The organiser could split the session, choose one primary role, or state that the event is an orientation rather than a skills workshop. The example is a constructed planning scene, not a reported customer result.<\/p>\n

What the policy evidence supports<\/h2>\n

It supports saying that Chinese education policy documents are treating AI capability, teacher development and education application as distinct planning concerns. It supports researching what a role-specific training offer would need to explain and what evidence a reviewer might request.<\/p>\n

It does not support saying that every teacher needs the expert\u2019s course. It does not establish an open procurement, a platform route, a standard curriculum, permission to use a particular policy phrase, or a learner outcome. It does not tell an overseas expert whether their examples, tools or assessment method fit a specific institution.<\/p>\n

This source boundary should appear in the brief before the curriculum is translated. Otherwise a broad policy sentence becomes an unspoken sales claim.<\/p>\n

Role-specific does not mean four separate businesses<\/h2>\n

An expert does not always need four complete products. A common conceptual core can remain shared, while the work task, examples and review artifact change by role. One course might have a short common foundation and four optional paths. Another might choose one role for a small test.<\/p>\n

The decision depends on what is being promised. If the promise is \u201cunderstand the language and limits of generative AI,\u201d a common orientation may be reasonable. If the promise is \u201cdesign a school implementation plan,\u201d the audience and artifact must be narrower. If the promise is \u201cbuild a classroom activity,\u201d a leader-focused policy lecture cannot be the only evidence.<\/p>\n

A counterexample: when a generic orientation is correct<\/h2>\n

There are situations where a mixed audience needs a shared orientation. A conference session may only aim to establish vocabulary, risks and questions for later work. It should be labelled as orientation, not as a role-specific capability programme.<\/p>\n

Likewise, a private internal briefing may not need a complete assessment. The point is not to force every event into a multi-track curriculum. It is to make the intended work and limits visible.<\/p>\n

The offer-review checklist<\/h2>\n

Before an overseas expert proposes AI teacher training for China, ask:<\/p>\n