An evidence-first field guide to AI use cases, workflows, tools and compliance, for sellers and operators. This book started with 34 public videos about using AI in ecommerce, from one-minute course ads to hour-long operator discussions. Each claim was checked against what the video actually showed, then against earnings calls, regulator documents, academic studies and platform policies. The result is less exciting than most AI business content, and more useful: some workflows are excellent, many income claims have nothing behind them, and several popular tactics now carry legal risk that nobody mentions. It covers where AI really pays off in a store (data and reporting, customer service, internal knowledge), six business routes compared by capital, skill and risk, a path from zero to first sale, twelve field cases, a prompt library, a tool map, getting found by AI shoppers, compliance and risk, and a 90-day plan. It ends with seven questions to ask of any AI money claim before you buy a course or a tool. Every claim carries a label (Shown, Claimed, Model or Simulated), so you can tell evidence from marketing. Research is current to 1 November 2026. It is general information, not legal, tax or financial advice, and it contains no affiliate links.

• The ten findings: what the evidence says about AI and ecommerce
• Where AI actually pays off, ranked by experienced brand operators
• Six business routes compared by capital, core skill and main risk
• Twelve field cases, a prompt library and a tool map
• Compliance and risk: fake testimonials, AI ads, duties and platform rules