Your code is a political act. Every line you ship facilitates something in the world: security, safety, abuse, or harm. For decades, developers have been ethical stewards of their work, but we never documented it; we passed wisdom through stories, mentorship, and code reviews. Now, AI generates code faster than you can review it and development teams are smaller. The ethical responsibility hasn't changed, but its weight has intensified; you don't have to write the harm to cause it.

In this talk, we'll examine how the ethics you've always practiced need sharper teeth. We'll look at what harm really means, from individual users to entire demographics. And we'll talk about why documenting your values, patterns, and guardrails isn't bureaucracy, it's how you collaborate ethically with AI. Because you can't surrender to the volume; the code you didn't review is code that still gets shipped.

Elevator Pitch
AI generates code faster than you can review it, and the ethics you've always practiced just got heavier.

Takeaways
Apply the harm lens at every level — individual users, demographics, and society
Push back on AI like a pair programmer — with alternatives and refined prompts, not just objections
Document the wisdom that used to live in code reviews — values, patterns, and guardrails your AI collaborator can inherit
Walk away when refusing the work is the ethical answer — pushback isn't always enough