Original work on agent behaviour, delegated authority, underwriting evidence, and the insurance systems an agent-run economy will need.
Three ISO endorsements now exclude generative AI from commercial general liability. What the forms say, where the carrier wording goes further, and which widely repeated claims about them do not survive a primary source.
What certification, a financial guarantee, and insurance around individual agents reveal about the first commercial interface for this market.
Why agent-native companies can change faster than annual insurance snapshots—and the evidence loop continuous underwriting may require.
A concept paper for one company-owned context packet that makes repeated market questions reusable without standardizing underwriting judgment.
An open proposal for a portable agent identity-and-evidence summary, mapped against the infrastructure already taking shape.
Orbital Compute tests how mission statements operate as meta-goals in AI-run economic simulations.
A manifesto for the continuous loop between frontier businesses, risk research, underwriting evidence, and new products for the agent economy.