Show HN: Humanitarian licensing and constitutional governance for AI agents

1 point

4 months ago

Yesterday I integrated the AOS Constitution (git-stamped Jan 1, 2026; https://aos-constitution.com) into a local-first agent (OpenClaw). The session later restarted with effectively zero retained context, so I asked the agent to review our humanitarian license with fresh eyes. It caught an overclaim ("uses automated monitoring" before we'd actually built monitoring). We fixed it in v1.0.1 the same day.

I'm Gene (founder of AOS). This is a constitutional governance toolkit + a humanitarian, source-available license for tool-using AI agents.

Repo: https://github.com/genesalvatore/aos-openclaw-constitutional License: https://github.com/genesalvatore/aos-openclaw-constitutional... License checker: https://aos-license-checker.com/

What it includes: - deterministic risk classification + policy evaluation (deny/confirm/allow) - signing/verification scaffolding + templates - humanitarian license with permanent downstream restrictions: - 40 prohibited categories (incl. absolute military prohibition, even "defensive") - copyleft-style inheritance (prevents license stripping) - commercial audit/self-cert requirements - patent separation (Section 5)

Controversy upfront: this is NOT OSI open source. It's source-available with explicit humanitarian values (no military use / weaponization). That's deliberate.

Why I'm posting: tool-using agents need governance before they're deployed at scale. We're trying to make governance concrete and auditable (not "prompt-based ethics").

Questions/criticism welcome — we're learning in public.

1 comment