Ask a school what its AI literacy programme costs and most Senior Leadership Teams (SLTs) can answer within a minute. Ask who will actually be running it in three years, once the launch budget is spent and the mandate has become routine, and the answer gets noticeably vaguer. That gap, between who delivers the first session and who owns the capability after it, is the real decision a school makes when it chooses a delivery model, whether or not the procurement conversation ever names it that plainly.
Most schools I talk to frame the choice as a shortlist of vendors, weighed on credentials, price, and how polished the sample lesson looks. That is a reasonable way to compare vendors against each other, but it skips a prior question: is a vendor even the right shape of answer for something a school will need to keep doing, in some form, for as long as the mandate exists.
What vendor delivery actually buys a school
A good vendor gives you speed and polish: a qualified trainer, a tested curriculum, and a launch that looks professional in front of the board or a parent evening within weeks rather than a term. For a school under calendar pressure ahead of 2027, that is not a small advantage, and I would not tell any school to turn it down on principle alone. But it is borrowed capability, not owned capability. The knowledge of how this particular group of students responded, which activities worked and which fell flat, how to slow down for a mixed-ability Year 8 class, stays with the trainer, not the school. It leaves when the contract ends, unless someone has deliberately arranged to write it down and keep it.
Before you sign with any AI literacy vendor, ask one direct question: what happens to this programme the year after the contract ends, if we don't renew? If the honest answer is "we'd have to start again," you are buying a one-off event, not something the school will own. Better to know that before you sign than after.
What changes when your own teachers deliver it
Teacher-delivered is slower to start, because you need to train the trainers before the first lesson can even happen, and that upfront cost is real. What you get back is a programme that builds on itself instead of starting from scratch each year: next year's version gets better because this year's teachers got better at teaching it, not because a vendor drew up a new contract. A teacher who delivers the lessons herself also knows the students, the curriculum, and the parents, in a way a visiting trainer working from a fixed slide deck never can. She can change course in the room when something isn't working, and a vendor's fixed programme simply isn't built to do that.
I had an interesting conversation with a fellow educator the other day about sustainable programmes in institutions. She quite rightly thought that it was in large part the leadership of an institution that made or broke such programmes, but that frequently the question of short and long-term costs were the deciding factor. "But monetary costs are just one part," I responded. "What's the real cost of starting a good programme and not being able to sustain it in the long term?"
That is the question a vendor-only model quietly avoids, because the vendor is not the one left paying the cost when a programme stalls after the invoice is settled.
Why this matters more across a multi-year mandate
A one-off compliance task can reasonably be handed entirely to an outside vendor. A multi-year responsibility can't, not without the costs adding up in the wrong direction. The 2027 mandate is not something a Malaysian international school does once and files away. It is an ongoing job that carries on every year after, through staff changes, curriculum updates, and new groups of students who arrive already using tools last year's programme never planned for. A model that starts from zero every time a contract ends or a vendor changes account managers is a much worse fit for that kind of ongoing job than a model where the knowledge stays with people who are still on staff next August or January.
None of this is an argument against ever using outside help. It is an argument for being clear about what that help is actually for. Outside expertise is well suited to designing the programme, training your trainers, and giving the kind of independent, vendor-neutral advice your own staff may not have yet. It is poorly suited to being the only way students actually get the lessons, term after term, once the design work is done.
Not sure which model fits your school?
A Discovery & Readiness session looks at your calendar, curriculum, and staff capacity before recommending a delivery model, rather than assuming teacher-delivered, vendor-delivered, or a mix of both is automatically right for you.
See How Discovery & Readiness WorksWhat This Means for You
Vendor training has a real role in getting a programme off the ground quickly, and I would not ask a school to slow down its 2027 preparation just to insist on teacher-delivered from day one. But the delivery model is a decision that adds up every year the mandate is in force, not a detail to sort out later. The question worth asking before you sign any contract is not only whether the vendor is good. It is whether the model itself leaves your school better equipped once the invoice is paid, or simply back where it started, waiting for the next one.