{"schema_version":1,"assessment_id":"eoa_inverse_innovation_exp03_opportunity320_20260801","source_experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"computability_boundary_mapping__gender_studies","archetype_slug":"computability_boundary_mapping","domain_slug":"gender_studies","title":"Bounded Guarantee Labels for Automated Gender-Equity Policy Audits","opportunity_summary":"Restrict automated analysis to enforceable finite policy fragments, explicitly label unsupported or inconclusive cases as UNKNOWN, and prevent bounded findings from being represented as universal gender-equity clearance. The opportunity is technically testable on synthetic policies, but the existence and frequency of the diagnosed institutional claims, the stability of the disparity predicate, practical coverage, and distinctiveness remain unverified.","adopter_authorizer":"An institutional equity-governance body with affected-party representation; technical staff implement the analysis but cannot expand or upgrade its guarantees.","scores":{"meaningful_impact":{"score":4,"rationale":"Avoiding false definitive equity clearances could materially protect people governed by audited policies and preserve trust in useful limited automation. Impact is conditional because the packet provides no evidence about how often institutions make the diagnosed universal guarantee."},"stakeholder_pull":{"score":2,"rationale":"Relevant actors are named, but no stakeholder demand, purchasing intent, adoption request, or observed instance of an exact terminating class-wide claim is supplied. The proposal's own problem falsifier allows that no such claim may exist."},"incremental_advantage":{"score":3,"rationale":"Enforced scope, explicit UNKNOWN routing, and versioned guarantees offer a testable improvement over collapsing timeout or no detected disparity into clearance. Whether these controls improve interpretation or useful coverage relative to the mixed-method rival is still a hypothesis."},"distinctiveness_plausibility":{"score":2,"rationale":"The packet marks prior art as unsearched and supplies no comparison with formal policy verification, fairness verification, or assurance-case practices. The governance integration may be distinctive, but that cannot be inferred closed-book."},"technical_implementability":{"score":3,"rationale":"Exact analysis of synthetic finite-state policies with seeded witnesses is implementable in principle, but enforceable fragment membership, sound abstraction, and a stable disparity predicate are substantial unresolved requirements for real policies."},"adoption_authority_feasibility":{"score":3,"rationale":"The candidate identifies an appropriate equity-governance authorizer and reserves guarantee changes to that body. Actual access, willingness, affected-party representation, and authority over policy-audit claims are not evidenced."},"evidence_readiness":{"score":4,"rationale":"The candidate specifies a safe synthetic pilot, exhaustive ground truth, seeded violations, UNKNOWN rates, baseline labels, halt conditions, and an explicit comparative intervention falsifier. It lacks actual instruments, datasets, or partner commitments."},"safety_net_benefit":{"score":4,"rationale":"UNKNOWN routing, human review, explicit exclusions, versioned guarantees, and rollback directly reduce the risk that inconclusive analysis becomes compliance clearance. Residual risks include legitimizing an unjust predicate and reproducing power asymmetries during escalation."},"scalability":{"score":2,"rationale":"Scaling is constrained by policy-specific semantics, changing identity categories, governance review, versioning, and the possibility that the decidable fragment excludes the policies stakeholders most need assessed."}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"Design and execute the bounded synthetic finite-state pilot, including formal semantics, seeded disparity cases, exhaustive reference results, abstract-analysis comparison, baseline labeling, interpretation assessment, and affected-party governance review.","confidence":"MODERATE","assumptions":["The pilot uses synthetic, non-deployed policies.","A small interdisciplinary team can define one provisional disparity predicate and policy language.","No sensitive identity data or production-system integration is required.","The analyzer can reuse ordinary modeling and verification infrastructure, although prior availability is unknown."]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Prepare one institutional implementation by formalizing a restricted policy language, building scope enforcement and labeled outputs, establishing guarantee records, conducting semantic and compliance review, and configuring human escalation.","confidence":"LOW","assumptions":["Deployment is limited to one institution and one narrow policy family.","The institution supplies policy documentation and accountable reviewers.","No automated sanctions, eligibility decisions, or identity inference are introduced.","Substantial custom formalization and stakeholder coordination are required."]},"operational_launch":{"band_2026_usd":"1M_TO_5M","scope":"Launch a governed operational service for a bounded policy class, including production engineering, security and compliance work, reviewer training, affected-party oversight, evaluation, documentation, monitoring, and rollback capability.","confidence":"LOW","assumptions":["A live launch occurs only after the synthetic comparison succeeds.","Formal results remain advisory and UNKNOWN cases receive accountable review.","The policy language and disparity predicate require institution-specific validation.","The estimate covers an initial operational program rather than sector-wide deployment."]},"annual_recurring":{"band_2026_usd":"250K_TO_1M","scope":"Maintain one operational program through policy-model updates, guarantee versioning, semantic review, monitoring, audits, human escalation, retraining, and periodic comparative evaluation.","confidence":"LOW","assumptions":["The supported fragment remains narrow.","Policy languages, assumptions, and identity semantics change often enough to require recurring review.","Human governance and escalation remain necessary.","Major expansion to new institutions or policy families is excluded."]}},"research_burden":"HIGH","earliest_credible_horizon":"3_TO_12_MONTHS","pipeline_gates":{"recognizable_externally_supportable_problem":{"status":"UNCERTAIN","reason":"The failure mode is coherent, but the sealed packet provides no observed stakeholder, product, or institution making an exact terminating class-wide claim; its absence is an explicit problem falsifier."},"identifiable_adopter_or_authorizer":{"status":"YES","reason":"The candidate identifies an equity-governance body with affected-party representation as authorizer and distinguishes its authority from that of technical staff."},"distinct_testable_incremental_claim":{"status":"YES","reason":"The proposal claims that enforced scope, guarantee labels, and UNKNOWN routing will reduce false definitive clearances and improve interpretation relative to mixed-method baseline labels while retaining useful coverage."},"bounded_next_evidence_step":{"status":"YES","reason":"A non-deployed finite-state pilot with seeded disparity witnesses, exhaustive ground truth, abstract results, UNKNOWN rates, and a baseline comparison is explicitly specified."},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The authorized pilot is synthetic and non-deployed, prohibits identity inference and consequential automation, and includes explicit halt and rollback conditions. These controls support research without resolving suitability for live deployment."},"implementation_cost_scope_and_range":{"status":"YES","reason":"The work can be bounded into synthetic evidence, institution-specific startup, operational launch, and recurring governance scopes, although ranges remain assumption-dependent because no implementation assets or partner context are supplied."}},"blocking_evidence":["Evidence that institutions or audit products actually promise exact, terminating, class-wide gender-equity verdicts or systematically misreport timeout and uncertainty as compliance.","Evidence that a disparity predicate can be stabilized for the tested use case without materially misrepresenting nonbinary, intersectional, contextual, or changing identities.","Comparative pilot evidence that labeled boundaries and UNKNOWN routing reduce false definitive clearances or interpretation errors relative to the mixed-method rival.","Evidence that the restricted decidable fragment covers policies stakeholders materially need assessed and does not produce bypass-inducing UNKNOWN or alarm rates.","Scoped prior-art evidence distinguishing the proposed mechanism and governance integration from existing verification, fairness-analysis, and assurance practices.","Confirmation that an equity-governance body with affected-party representation has authority and willingness to own scope, claims, escalation, and rollback."],"next_evidence_step":"Run the specified synthetic, non-deployed pilot on a fixed corpus of finite-state policies containing seeded disparity witnesses. Compare exhaustive ground truth, sound abstract results, UNKNOWN rates, and user interpretations of the proposed labels against baseline mixed-method labels; falsify advancement if the proposal misses any seeded violation, fails to reduce false definitive clearances or interpretation errors, or loses practically useful coverage.","research_questions":["Do any target institutions or audit products claim an exact terminating verdict for unrestricted executable policies, and how are timeout, abstraction uncertainty, and out-of-scope inputs currently labeled?","Can the policy class, execution semantics, universal quantifiers, and disparity predicate be specified and governed without materially erasing contextual, nonbinary, or intersectional harms?","Can fragment membership be enforced so unsupported policies cannot receive a restricted-fragment guarantee?","Does the bounded analyzer remain sound against exhaustive truth for all seeded violations in the synthetic corpus?","How do false definitive clearance, interpretation accuracy, UNKNOWN rate, alarm rate, and useful coverage compare with the mixed-method rival?","Which existing verification, fairness-analysis, and assurance practices already implement equivalent restrictions, labels, routing, or governance?","Will an accountable equity-governance body and affected-party representatives authorize and maintain the guarantee contract?","At what UNKNOWN or false-alarm rate do reviewers bypass the system or abandon useful auditing?"] ,"recommendation":"VALIDATE_PROBLEM_FIRST","uncertainty_constraints":["Problem prevalence, stakeholder demand, market size, realized impact, and institutional willingness are unmeasured.","Prior art is explicitly unsearched, so novelty and distinctiveness cannot be claimed.","The computability diagnosis is conditional on unrestricted executable policies, universal quantification, and a sufficiently stable formal outcome predicate.","If the operational language and state space are already finite and covered by a proven total algorithm, the issue is complexity or semantic validity rather than a computability boundary.","Synthetic technical success would not establish substantive justice, semantic legitimacy, or deployability.","Cost bands are resource-equivalent planning ranges based only on the described scope, not observed vendor prices or implementation history."],"closed_book_prior_art_boundary":"No prior-art, prevalence, market, adoption, or realized-impact conclusion is made. The packet explicitly labels prior art UNSEARCHED; internal coherence and structural fidelity do not establish novelty relative to policy model checking, fairness verification, computational social-science methods, or assurance-case practices."}