That's actually what I'm trying to avoid: using trust assumptions. This approach relies on alignment of incentives instead...