Level 3 taught a team to run AI-assisted plant work as an operation across parts, lines, shifts and customers. Level 4 asks the people who answer for the whole company to take responsibility for AI: owners, presidents and general managers, the leaders of operations, quality, engineering, supply chain and EHS, the export-control officer where one exists, and the IT lead who has been handed the AI question. Every module ends with something a customer auditor, a registrar, an insurer, a prime contractor, a regulator or the plant's own workforce could pick up and test.
The first three modules place accountability and record it. Module one explains why responsibility for AI scatters across operations, quality, engineering, EHS, IT and the owner until nobody holds it, places it with named people for safety, product conformity, controlled technical data and company information, and names what no tool, integrator or vendor can take off the company. Module two writes a governance programme sized to a plant with no legal department: a permitted-use inventory, data classes, decision rights, stop authority, supervision and a review cycle, placed beside the quality management system without pretending a quality standard requires it. Module three builds the decision record and the evidence pack, and keeps both alive when the people who made the decisions move on.
Modules four to six map what binds the plant. Module four builds an obligation map with law, law arriving through a contract, voluntary standards and certification schemes in separate columns, covering workplace safety, environmental reporting, product hazard reports, origin claims and state and foreign AI statutes, with a method for deciding what a change means. Module five makes export control a governance function: who holds the authority, the screening gate before any file reaches any service, technology control across drives, AI tools and plant access, and the record that proves screening happened. Module six reads a prime's defence flowdown clause by clause, states the certification programme as real and dated without guessing at what is not yet known, and asks whether controlled unclassified information may reach a cloud AI service at all.
Modules seven to ten cover what the company buys. Module seven assesses a supplier on where plant data goes, how long it stays, whether it trains a model and how it is deleted, tells an answer from an evasion, and shows why the file you keep matters to trade-secret protection. Module eight reads a performance claim for what was measured, by whom and on whose data, explains why a public benchmark cannot speak for your parts, and designs an acceptance test on the plant's own material. Module nine turns due diligence into enforceable terms for data use, model change, incident notice, deletion, exit and acquisition. Module ten deals with AI that arrives inside a purchased machine, the remote access that comes with it, updates that change it, and a learned function near a safety function you did not specify.
Modules eleven and twelve handle security and fraud. Module eleven places every AI system on one asset inventory across OT and IT, with identity, segmentation, monitoring, supplier oversight, media sanitisation and incident response, using voluntary federal frameworks as frames rather than as law. Module twelve assumes a familiar voice or a plausible email proves nothing, writes call-back verification for bank changes, expedites and customer approvals before it is needed, plans the first hour after a suspected impersonation, and sets out the state safe-harbour and breach duties that may follow.
Modules thirteen and fourteen close with leadership. Module thirteen takes experienced people's objections seriously, gives a straight answer on jobs, runs a pilot on the plant's own material and measures adoption without a borrowed figure. Module fourteen sequences the programme from the lowest-risk use outward, states costs with risks avoided and the plant's own baseline rather than a published return, and writes the report the owners decide on. The capstone is an AI Governance Programme, Scenario Defence and Ownership Report built on a fictional plant. The level ships with a printable workbook and ten templates, and the examination draws forty scenario questions from a reviewed bank.
Everything here is professional education. It is not legal, engineering, quality, safety, cybersecurity or export-control advice, and it does not replace the employer's safety programme, a licensed professional engineer where a state requires one, the person your quality system names, an empowered customer representative, your export authority, legal counsel, or the law that applies to your plant. Completing the level earns an independent educational certificate issued by AI Coalition Network with a public verification page. It is not an engineering licence, a quality-system registration, a safety qualification, a cybersecurity certification or a regulatory approval, it carries no professional development hours, and it qualifies no one to audit, assess or sign anything under any standard, scheme or contract.