GenZAgents for ecommerce — AI-assisted product, content, and pricing audit

UK CMA + EU UCPD + US FTC all require commercial communications to be truthful and not misleading. AI-generated product descriptions, comparisons, reviews — if misleading, you're liable. The audit trail showing supervisor review + AI model + change log is your defensive position. Without receipts: you're relying on engineering memory. With: documented evidence.

AI-driven dynamic pricing: pricing engine output becomes a receipt. The audit trail tracks per-customer pricing decisions + the AI model + the inputs. For the EU's P2B Regulation 2019/1150 (platform-to-business transparency) and emerging AI-pricing transparency rules, this audit trail is the compliance evidence.

FTC has explicitly warned against undisclosed AI-generated reviews. UK ASA same. Receipt-tagging of AI-assisted content lets you build operational controls: "no review tagged ai_generated_review=true gets published without human review and disclosure label". The audit trail makes ASA / FTC compliance verifiable rather than hoped-for.

AI personalisation hits GDPR profiling rules. The receipt feed lets you respond to subject access requests with "here's the AI activity touching your profile data". GDPR Article 22 right against automated decisions is supportable with documented evidence of human review.

Merchandiser uses Claude to generate product descriptions at scale. Each generation is a receipt — merchandiser, product SKU, model, copy (digest). The audit trail shows the human review chain. If a customer complains the description is misleading, the receipt anchors the investigation.

AI chatbot handles initial customer service. Each conversation is a receipt — customer ID, chatbot session, model, escalation status, supervisor (if escalated). The audit trail satisfies both consumer-rights documentation (showing how customer queries were handled) and AI Act §50 transparency (AI vs human disclosure).