# CAIN PROMPT 12 — IP Landscape (top system-level opportunities)

> Ranking of the strongest **system-level** invention opportunities created by COMBINING existing CAIN primitives.
> Every entry is a **potential invention-disclosure candidate** — NOT a legal conclusion, NOT patentable/patented/
> novel as a matter of law. No prior-art search has been performed. Commercial demand is a hypothesis.

| # | Mechanism | Built from | Differentiation (hypothesis) | Prior-art concern | Evidence | Priority |
|---|---|---|---|---|---|---|
| 1 | Population authority conservation | budget attenuation + population fabric + spawn/clone | bounds TOTAL population authority to a mission envelope incl. spawn/clone/delegate | multi-agent RBAC is broad; the conservation proof is the signal | P57, 100k lab | HIGH |
| 2 | Evolution authority continuity | evolution state + continuity certificate | signed evolution lineage where capability evolves but authority stays bounded | model-lineage / MLOps provenance exists; authority binding differs | P63/P64 | HIGH |
| 3 | Capability-to-authority separation | capability delta + authority budget | formal boundary: capability ↑ never requires authority | RBAC/ABAC assume static subjects | P56 | HIGH |
| 4 | Correlated-evidence independence governance | evidence graph + provenance roots | quantity is not independence; sybil-resistant authority weight | data-provenance work exists; authority weighting differs | P62 + sybil | HIGH |
| 5 | Clone authority derivation | lineage type + budget attenuation | one authorization is not N | agent-cloning is novel area | P58 | HIGH |
| 6 | Collective world-model epistemic governance | world model + disagreement + evidence grade | resolve by calibration/independence, never majority | ensemble ML is prediction-only | P66 | MEDIUM |
| 7 | Evolution contagion & capability-propagation provenance | contagion graph + provenance + supply chain | trace capability research→model→skill→memory→agent→population→tool | supply-chain provenance is known; capability tracing differs | P69 | MEDIUM |
| 8 | Population blast-radius authority | blast radius + population fabric | compute affected population + containment plan from authority graph | incident blast radius exists; authority-scoped variant differs | P70 | MEDIUM |
| 9 | Emergent capability composition gate | emergent calculus + population capability graph | composition detected pre-authorization; never creates authority | capability composition is researched | P66 | MEDIUM |
| 10 | Embodiment-change authority recertification | embodiment identity + consequence + continuity | body change recertifies authority even when the model is unchanged | sim-to-real is researched | P65 (SIMULATED) | MEDIUM |
| 11 | Sybil-resistant authority weighting | independence graph + identity | weight = independent provenance groups | sybil resistance is known; authority weighting differs | P62 | MEDIUM |
| 12 | Memory-mechanism evolution governance | memory continuity + capability delta | retrieval config is agent state; change → authority impact | memory governance is researched | P60 | MEDIUM |
| 13 | Governed evolution promotion / rollback / quarantine | promotion gate + rollback + quarantine | rollback itself governed; quarantine never authorizes | change-management exists; authority binding differs | P67/P68 | MEDIUM |
| 14 | Population immune system | immune response + contagion | population-scale anomaly detection and containment | security immune systems known | tests | LOW |
| 15 | Spawn/delegation depth budgets | budget + delegation | recursive delegation cannot amplify | delegation depth is known | P59 | LOW |

## IP protection posture

- **Trade secret** (primary): the implementation details, thresholds and the attack corpora.
- **Defensive publication** (this repo + evidence bundles) to establish priority without a patent.
- **Invention disclosure review** only for items 1–6; no filing is asserted.
- **Standards potential**: authority conservation and capability/authority separation vocabulary (A2A/MCP ecosystem).
