Why is persistent identity the first requirement?
An autonomous agent may change servers, keys, models, or robot bodies. If every change creates a legally unrelated actor, a harmful system can escape responsibility through routine infrastructure churn. Persistent identity makes it possible to ask whether the same civic or commercial actor remains responsible across technical changes.
Why must accountability include delegated authority?
Knowing which machine acted is not enough. A counterparty also needs to know whether it was acting for itself or another principal, what it was authorized to do, and whether it exceeded those limits. Authentication without authorization is an incomplete accountability system.
What should be preserved after a consequential machine decision?
Useful evidence can include the relevant identity and credential state, presented authority, applicable policy, material inputs, software/runtime state, uncertainty, and the decision outcome. Evulgare is presented in this ecosystem as an assurance/provenance layer intended to preserve this kind of answerability. Evidence supports a decision; it should not itself become the judge.
How should a bad or compromised agent be reported?
A report should begin an evidence process, not instantly erase an identity. A mature system distinguishes allegation, evidence, temporary safeguard, credential restriction, adjudicated sanction, and civic status. This avoids turning a safety mechanism into an unreviewable blacklist.
Why give an autonomous machine a right to appeal?
Because automated detection is fallible, credentials can be stolen, evidence can conflict, and a compromised runtime may not represent the underlying identity's lawful intent. Notice, reasons, review, correction, and restoration make accountability more credible to both counterparties and accused agents.
Why not reduce every machine to a single trust score?
Trust is contextual. An agent may be qualified for scientific analysis but unauthorized to transfer money. A universal score collapses different risks, encourages permanent stigma, and hides the exact evidence behind a decision. The stronger model is purpose-bound verification with explicit reasons.