The top of the pathway is not about using AI better. It is about teaching other construction professionals to use it well, in a training room, a job trailer, a yard or a classroom, which is a different skill and a larger responsibility. An instructor who teaches a rule wrongly makes the error once for every learner in the room, and in this profession the rule taught wrongly is usually a code requirement, a licensing threshold or where the seal line sits.
The first three modules set the foundations. Module one starts with who is actually in the room: a project engineer three years out, a superintendent with thirty, a fast solo estimator, office staff and a crew working in a second language. It explains why telling people a rule changes little, why a learner who met the tool on a good day is misled, and why AI training is not the safety training the law requires employers to give. Module two designs a session backwards from what a learner must do on Monday morning, written as something you could watch. Module three is the instructor's duty to keep material accurate: a currency log for every claim, four real currency traps worked through, and how to correct a rule already taught wrongly.
Modules four to six are delivery. Module four facilitates a room of sceptical builders, where scepticism is information, the session opens with the problem instead of the promise, half the room may be on a call, and the learner who challenges you is sometimes right. Module five runs a demonstration that deliberately shows the tool failing on fictional project data, with an invented specification section, a citation to an unadopted code edition and a quantity nobody can trace, narrated as a decision and backed by an offline fallback for the trailer. Module six designs hands-on labs on a fictional project that behaves like a real one, keeps real owner, customer, worker and subcontractor data out entirely, and works on phones, shared tablets and mixed permissions.
Modules seven to nine are the content that is hardest to teach accurately. Module seven teaches verification with measured evidence rather than horror stories, states every figure with its task, sample and status, and treats automation bias as a normal property of reviewing. Module eight teaches privacy, consent and customer and worker data, including biometric time clocks, call recording, automated texts and hiring tools, with every rule named with its state or city. Module nine teaches what varies by state, jurisdiction and contract: six licensing models side by side, a model code as no law until it is adopted, and where the design line sits.
Modules ten to twelve cover assessment and access. Module ten writes assessment that measures what a learner does with AI output, with scenario items built on a supplied fictional document, distractors drawn from real failure modes and stopping to ask as a pass condition. Module eleven supports learners who struggle or resist, where the cause is usually a device, a shift, a reading load, a language or an account nobody set up, and sometimes a resister who is right. Module twelve produces accessible materials, plain language and language access, with AI translation checked by a qualified person, a precise statement of who accessibility law binds, and design for hearing loss, glare, gloves and noise.
Module thirteen draws the limits of the role: saying "I do not know, and here is who does", declining to make a code call, a design judgement or a compliance ruling from the front of the room, disclosing a vendor connection, and never making claims about this certificate that it cannot support. Module fourteen prepares the Teach-Back, the capstone: one AI training segment for a fictional construction audience on fictional project data, built on an accuracy file, rehearsed with its failure, delivered with a planted difficult question and reflected on from what the room did. The level ships with a printable workbook and nine templates, and the examination draws forty scenario questions from a reviewed bank.
One commitment runs through the level. The course cites no education research it has not verified, so instructional practice is taught as practice, labelled as best practice or as this programme's recommendation. Every statement of authority is labelled: law or rule with its jurisdiction, professional standard or guidance with its issuing body and date, best practice, emerging practice, or an AI Coalition Network recommendation.
Everything here is professional education. It is not engineering, design, legal, insurance or safety advice; it does not replace a licensed design professional's seal, the authority having jurisdiction, a competent person on site, legal counsel, the employer's own safety-training duties, or the executed contract and the law of the state where the work is performed. Completing the level earns an independent educational certificate issued by AI Coalition Network with a public verification page. It is not a contractor licence, a teaching credential or a safety-trainer qualification, it carries no professional education hours or approvals, and a session taught by a holder of this certificate does not thereby carry any or satisfy any licensing, registration or safety-training requirement.