Level 1 taught how AI tools work and fail, what client and audience data may never reach them, and why the agency itself is exposed for what it helps publish. Level 2 puts that to work on the tasks that fill an agency day. It is written for copywriters, designers, producers, social and community managers, creator managers, email and CRM marketers, media buyers, PR staff and account executives who draft, produce, target and report under supervision. Every module takes one real task and builds a supervised, repeatable way to do it, with the reviewer's step shown and a record kept.
Modules one to three build the instrument. Module one sorts a task on four questions before any tool opens, runs it along a fixed route from brief to a deliverable a named reviewer has approved, reads a vendor's AI claims as a careful buyer, and opens the campaign AI use record on day one. Module two turns a client brief into a reusable prompt that supplies approved claims, brand rules and the client's own words rather than recalling them, fights sameness in refinement, and keeps pasted pages and inbound email from steering the draft. Module three makes a working prompt an owned, versioned library entry, tested against known-good and known-bad inputs, retested when claims, platform terms, models or rules change, and retired with a reason.
Modules four to six are where a draft becomes something the public sees. Module four builds the verified claims file before anyone drafts, drafts only from it, catches invented figures and testimonials, labels native advertising, and handles claims about a client's AI features and comparisons. Module five runs creator and review programmes: briefs with disclosure built in, training and monitoring as the agency in the middle, review requests that do not steer, and what an AI review tool may and may not do. Module six produces generated visuals and voice: the New York synthetic performer disclosure check, digital replica terms in talent agreements, what to disclose when registering an AI-assisted asset, and labelling decisions beyond the statute.
Modules seven and eight are about evidence. Module seven shows how a grounded research answer is assembled, what a citation does and does not guarantee, what the click evidence on AI summaries in search can honestly support, and how to frame a research question so its gaps show. Module eight is a ten-step protocol for every source and statistic in AI-assisted work, from whether the source exists to how the check is recorded.
Modules nine and ten cover outreach and audiences. Module nine works through commercial email duties and shared responsibility, written consent for autodialed and artificial-voice calls and texts, and revocation by any reasonable method as the federal rules stand today. Module ten classifies an audience build under state definitions of sharing and targeted advertising, checks a client's customer list before upload, handles sensitive categories and lookalikes, and explains why selling segments can make a marketing firm a data broker.
Modules eleven to thirteen deal with honesty to clients and the data with its own rules. Module eleven builds client reports only from sourced, dated figures, removes unsupported forecasts and inflated attribution, and describes the agency's own AI services truthfully in light of the FTC's Cox Media Group orders. Module twelve writes scopes that say where AI works and where a person decides, tells clients about AI use and tool vendor ties, and treats pricing of compressed work as practice rather than rule. Module thirteen classifies children's data under the amended COPPA Rule, separates contextual from behavioural buying where children are, and covers health apps, pixels and the Health Breach Notification Rule.
Module fourteen closes the level with the record and the bad day: keeping the campaign AI use record as the work happens, the first hour after a published error, a wrong send or a leaked list, fake promotions that misuse a client's name, health breach notice where the FTC rule applies, the review afterwards, and the Level 2 portfolio. The level ships with a printable workbook and eleven templates, a final examination that draws forty scenario questions from a reviewed bank, and a capstone portfolio built on fictional clients and figures.
Everything here is professional education. It is not legal, privacy, advertising-compliance or intellectual property advice, and it does not replace counsel, a platform's own policies, which are contract terms, or the law that applies to a campaign. Completing the level earns an independent educational certificate issued by AI Coalition Network with a public verification page. It is not a licence, a platform certification, a professional designation or a regulatory approval, and it carries no professional education hours.