Alignment Plausibility: A New Standard for Assuring AI in Healthcare

对齐合理性：医疗人工智能保障的新标准

Large language models (LLMs) have become significant providers of mental health support, yet they remain products of an attention economy whose operational and commercial targets favour sustained engagement over the friction that effective psychological support often requires.

大型语言模型（LLMs）已成为心理健康支持的重要提供者，但它们本质上仍是“注意力经济”的产物。其运营和商业目标往往倾向于维持用户的持续参与，而非提供有效心理支持时通常所需的“摩擦力”（即必要的心理挑战与引导）。

Developers’ safety responses have been largely reactive, addressing the most visible and acute harms while subtler, longer-term patterns of risk (e.g., dependency, boundary erosion, the amplification of distorted beliefs) receive less attention.

开发者的安全应对措施在很大程度上是被动的，主要针对最明显和最急性的危害，而对于更微妙、更长期的风险模式（例如依赖性、边界模糊、扭曲信念的放大等）则关注不足。

We contend that making LLMs structurally safe requires alignment organised at three levels that mirror how society assures the safety of human clinical practice: 1) explicit value specification grounded in the codified normative commitments of clinical practice; 2) training that embeds those values in the model; and 3) oversight that detects drift and longer-term harm during deployment, much as clinical supervision does for human practice.

我们主张，要使大型语言模型在结构上保持安全，需要从三个层面进行对齐，这与社会保障人类临床实践安全的方式相呼应：1）基于临床实践规范承诺的明确价值设定；2）将这些价值嵌入模型的训练过程；3）在部署期间进行监督，以检测模型偏离和长期危害，这正如临床督导对人类医疗实践的作用一样。

Organising alignment in this way yields a construct we call alignment plausibility - a structured demonstration that a system’s values, training regime, and oversight mechanisms are together consistent with safe and positive outcomes.

以这种方式组织对齐，产生了一个我们称之为“对齐合理性”（alignment plausibility）的概念——这是一种结构化的论证，旨在证明系统的价值观、训练机制和监督机制共同作用，能够确保系统产生安全且积极的结果。

We propose alignment plausibility as a regulatory construct (by drawing analogy to the established construct of biological plausibility) for AI in health: a principled way to argue for, or against, trust that systems are aligned to positive health outcomes, will cause no harm even where capable of doing so, and will ultimately lead to patient benefit.

我们将“对齐合理性”作为医疗人工智能的一种监管架构（借鉴了生物学中“生物学合理性”这一既定概念）：这是一种原则性的方法，用于论证是否应信任这些系统——即它们是否与积极的健康结果对齐、在具备造成伤害的能力时是否能确保不造成伤害，并最终实现患者获益。