Conformance as a Distribution: Delegation Contracts, Error Budgets, and the Allocation of Evaluation Authority in Use-Based AI Regulation

Computer Science > Artificial Intelligence arXiv:2610.33112 (cs) [Submitted on 23 Oct 2026] Title: Conformance as a Distribution: Delegation Contracts, Error Budgets, and the Allocation of Evaluation Authority in Use-Based AI Regulation Authors: John deVadoss.

计算机科学 > 人工智能 arXiv:2610.33112 (cs) [提交于 2026 年 10 月 23 日] 标题：作为分布的一致性：委托合同、错误预算以及基于用途的 AI 监管中评估权限的分配 作者：John deVadoss。

Abstract: Use-based AI regulation assigns obligations by deployment risk but lacks an evidentiary object to carry them. A benchmark score is a property of a model; conformance of a deployment is a failure rate under a declared distribution of contexts, bounded by a scope and subject to expiry.

摘要：基于用途的 AI 监管根据部署风险分配义务，但缺乏承载这些义务的证据对象。基准分数是模型的一种属性；部署的一致性是在声明的上下文分布下，受范围限制且有有效期的故障率。

We take the delegation contract of earlier work (frozen configuration, scoped claim, executable revision rule) and study it under strategic play between a deployer, an auditor, and an enforcement point. We ask whether a declared context mixture can be made truthful; whether an audit can be unbiased against a system that detects evaluation traffic; whether a certification error budget can sustain a long-lived deployment; whether an aggregate audit can support a per-subgroup guarantee; and whether an incident metric that counts system failures but not fallback costs ranks arrangements correctly.

我们采用了早期研究中的委托合同（冻结配置、范围声明、可执行修订规则），并在部署者、审计员和执行点之间的战略博弈下对其进行了研究。我们探讨了：声明的上下文混合能否做到真实；针对检测评估流量的系统，审计能否做到无偏；认证错误预算能否维持长期部署；聚合审计能否支持针对子组的保证；以及仅计算系统故障而不计算回退成本的事件指标能否正确地对安排进行排序。

For each we give a mechanism and a proposition: a proper-scoring penalty on the declared mixture bounds the excess failure of a best-responding deployer by $O(\lambda^{-1/2})$ in the penalty weight; a commit-reveal audit drawn from live traffic is unbiased by construction and grades the evaluator with the same sample; a halving error budget drives certification coverage below $(1-p)^n$ while $\alpha$-investing sustains it under a marginal false-certification guarantee; a per-subgroup guarantee costs audit size scaling as $1/m$ in subgroup mass; and any failure-only incident metric prefers withdrawal at every fallback cost, even when total loss ranks the arrangements oppositely.

针对每一项，我们都给出了机制和命题：对声明混合的适当评分惩罚将最佳响应部署者的超额故障限制在惩罚权重的 $O(\lambda^{-1/2})$ 以内；从实时流量中提取的提交-揭示审计在构建上是无偏的，并使用相同的样本对评估者进行评分；减半错误预算将认证覆盖率驱动至 $(1-p)^n$ 以下，而 $\alpha$-投资在边际错误认证保证下维持该覆盖率；子组保证的成本随子组规模的 $1/m$ 缩放；任何仅关注故障的事件指标在所有回退成本下都倾向于撤回，即使总损失对这些安排的排序截然相反。

Five preregistered synthetic studies test each claim; no language model or production system is evaluated. Enforcement sits at distribution, capability-access, monetization and risk-transfer choke points, expiry of the conformance record is a standard of care, and conformance-as-a-rate is set against the EU AI Act’s static conformity assessment.

五项预注册的合成研究测试了每一项主张；未评估任何语言模型或生产系统。执行位于分发、能力访问、货币化和风险转移的瓶颈处，一致性记录的过期是一种注意标准，而“作为速率的一致性”则是针对欧盟《人工智能法案》中静态合格评定的对比。