引擎盖之下：Uber 负责任 AI 的规模化

April 2, 2026

Under the Hood: Scaling Responsible AI at Uber

幕后：Scaling Responsible AI at Uber

April 2, 2026

2026年4月2日

Under the Hood: Scaling Responsible AI at Uber

幕后：Uber 扩展负责任 AI

April 2, 2026

Under the Hood: Scaling Responsible AI at Uber

Under the Hood：Scaling Responsible AI at Uber

Introduction

引言

As AI and machine learning become even more central to critical products and services—including at Uber—companies must understand how their models work and ensure they’re governed responsibly. At Uber, where AI is developed across many platforms and teams, we launched a company-wide Responsible AI program to bring visibility, explainability, and governance to models. Through a centralized Model Catalog, automated tooling, feature-importance explainability, early compliance checks, and broad employee education, we’ve built a scalable foundation for responsible AI innovation across the company.

随着 AI 和机器学习在关键产品和服务中变得更加核心——包括在 Uber——公司必须了解其模型如何工作，并确保它们得到负责任的治理。在 Uber，AI 由多个平台和团队开发，我们启动了一个公司范围的 Responsible AI 计划，为模型带来可见性、可解释性和治理。通过集中的 Model Catalog、自动化工具、特征重要性可解释性、早期合规检查以及广泛的员工教育，我们为公司范围内的负责任 AI 创新构建了一个可扩展的基础。

The Situation: AI Governance is Essential For Scale

现状：AI治理对于规模化至关重要

As AI systems become integral to our daily lives, ensuring they perform reliably and fairly is essential. People rightly expect the systems they interact with to be developed and governed responsibly.

随着 AI 系统成为我们日常生活不可或缺的一部分，确保它们可靠且公平地运行至关重要。人们有理由期望他们互动的系统以负责任的方式开发和治理。

To meet these standards of stewardship, Uber has implemented a governance framework that provides deep visibility and accountability for models across the Uber platform. This framework provides a clear understanding of what Uber’s models do, how they work, and how they’re governed end-to-end throughout the development and deployment process.

为了达到这些管理标准，Uber 实施了一个治理框架，为 Uber 平台上的模型提供深入的可见性和问责制。该框架提供了对 Uber 的模型做什么、如何工作以及在开发和部署过程中如何端到端治理的清晰理解。

The Complication: AI is Everywhere

难题：AI 无处不在

AI is deeply embedded across Uber, so effective governance must extend ac...