TWC: Small: A Choice Architecture for Mobile Privacy and Security

移动设备（例如，智能手机和平板电脑）允许用户执行丰富的第三方应用程序，这些应用程序能够广泛使用设备硬件和个人数据。这会带来安全风险，因为应用程序可能会执行不希望的操作，例如删除数据、损坏硬件，甚至直接引起用户话费的增加。移动设备也会带来隐私风险，因为它们存储了敏感的个人信息，这些信息可能会被不恰当地访问和共享。授权用户决定如何访问其移动设备上的资源（即“授予许可”）是移动计算未来的一个重要挑战。我们的研究表明，现有的机制是无效的：用户经常授予权限，因为他们要么不理解这些权限，要么习惯了这些权限，要么觉得别无选择。本研究项目旨在确定和研究这些问题的潜在解决方案。该项目开发了一种以用户为中心的移动设备许可请求方法。该项目正在进行人体实验，以设计和验证更有效的机制，以规范对隐私或安全敏感的行为。研究议程包括尽可能用受保护的部件（即“可信UI”）或审计机制代替安全警告，从而最大限度地减少对安全警告的习惯；改进安全警告的设计，因为其他许可授予机制有时是不合适的；并将这些机制集成到一个平台中，通过采用以用户为中心的方法授予权限来提高系统安全性。如果成功，这个项目将有助于为未来几代移动设备奠定一个安全的基础。

Mobile devices (e.g., smartphones and tablets) allow users to execute rich third-party applications that are capable of making extensive use of device hardware and personal data. This poses security risks, as applications may perform undesirable operations such as deleting data, damaging hardware, or even directly incurring charges on the user's phone bill. Mobile devices also pose privacy risks, as they store sensitive personal information that may be accessed and shared inappropriately.Empowering users to decide how resources on their mobile devices are accessed (i.e., "granting permission") is an important challenge for the future of mobile computing. Our research has shown that existing mechanisms are ineffective: users frequently grant permissions because they either do not understand them, are habituated to them, or feel that they have no other choice. This research project aims to identify and study potential solutions to these problems.This project develops a user-centered approach to mobile device permission requests. The project is conducting human-subjects experiments to design and validate more effective mechanisms for regulating privacy- or security-sensitive actions. The research agenda involves minimizing habituation to security warnings by substituting them with protected widgets (i.e., "trusted UI") or audit mechanisms, when possible; improving the design of security warnings, because alternative permission-granting mechanisms are sometimes inappropriate; and integrating these mechanisms into a platform that improves system security by taking a user-centered approach to granting permissions. If successful, this project could help develop a secure foundation for future generations of mobile devices.