Automation ethics notes we give every cohort on day zero

Automation changes how accountability shows up. We ask learners to name the human owner for every workflow before they enable triggers, even in sandboxes. We also ban dark patterns in class projects: no fake urgency SMS, no disguised marketing automations, no scraping personal inboxes without documented consent. Breaking those rules fails the capstone regardless of technical cleverness. When workflows touch protected attributes, we require a plain-language notice in the README about who approved the data use and when that approval expires. It is not legal advice, but it trains the reflex to involve the right stakeholders. These notes are non-negotiable. They keep our alumni employable in organizations that already worry about reputational risk from sloppy bots.