Answers are grounded in this paper and cite their sources.
Theoretical analysis reveals algebraic limits of defensive deception against generative adversaries, demonstrating that layered shields preserve mission-critical secrets despite configuration leaks.
How much of a deceptive guarantee depends on the adversary's ignorance of the configuration, and how much survives full knowledge of it? We answer for the class of knowledge-monotone deceptive guarantees by decomposing each one, exactly and algebraically, into a disclosure value surviving full knowledge of the realized configuration and a configuration-carried value vanishing at that knowledge (Proposition 3.1). The interactive experiment is a conguration read through adaptive adversary policies and transcripts, evaluated by typed functionals in two registers defensive advantages and adversarial risks. Resistance is obtained not by re-deriving an information-theoretic floor but by instrumenting the three defensive anchors of [Rouxel 2026c, Corollary 5.1]: a residual-separation anchor on the defender's observation channel (Proposition 4.3, an instance of [Rouxel 2026c, Proposition 5.1a]), a role-reversed secret shield on the adversary's acquisition channel (Proposition 4.4, an instance of [Rouxel 2026c, Proposition 5.1b]), and a reachability-exclusion anchor relative to the mission's compromise set (Proposition 4.8, an instance of [Rouxel 2026c, Proposition 5.1c]); suciency is proved, completeness is not claimed. The central result is a well-founded shield recursion (Theorem 6.4): a finite meta-tower of secret shields terminating on a separation or exclusion anchor bounds the acquisition of the mission-critical level-zero secret by a sum of per-level min-entropy terms, composed in guessing probability rather than in conditional entropy, a compromise bound following under mission-criticality. Attraction is re-founded on a relative base-rate bound (Proposition 5.4); the Kerckhoffs criterion is stated per anchor with a conjectural converse under a uniform capability floor (Corollary 7.1, Conjecture 7.2); the alternative paradigms reduce conditionally (Propositions 8.2 and 8.3). No result claims necessity, exhaustivity, or the discharge of any conjecture: sufficiency is proved anchor-wise, necessity is conjectural through [Rouxel 2026c, Conjecture 5.2].
No takes yet. Share an insight, caveat, or question.
Franck Rouxel (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: