Zero-click attacks require no user interaction and typically exploit zero-day (i.e., unpatched) vulnerabilities in instant chat applications (such as WhatsApp and iMessage) to gain root access to the victim's smartphone and exfiltrate sensitive data. In this paper, we report our experiences in attempting to secure smartphones against zero-click attacks. We approached the problem by first enumerating several properties we believed were necessary to prevent zero-click attacks against smartphones. Then, we created a security design that satisfies all the identified properties, and attempted to build it using off-the-shelf components. Our key idea was to shift the attack surface from the user's smartphone to a sandboxed virtual smartphone ecosystem where each chat application runs in isolation. Our performance and usability evaluations of the system we built highlighted several shortcomings and the fundamental challenges in securing modern smartphones against zero-click attacks. In this experience report, we discuss the lessons we learned, and share insights on the missing components necessary to achieve foolproof security against zero-click attacks for modern mobile devices.
翻译:零点点击攻击不需要用户互动,通常会利用即时聊天应用程序(例如“WhessApp”和iMessage”)中的零天(即未喷射的)弱点获取受害者智能手机的根接入和排泄敏感数据。 在本文中,我们报告了我们试图确保智能手机不受无点击袭击的经验。 我们首先列举一些我们认为防止对智能手机零点击袭击所必需的属性, 来解决这个问题。 然后, 我们设计了一个安全设计, 满足所有识别属性, 并试图使用现成组件来建造它。 我们的关键想法是将攻击表面从用户智能手机转移到一个每个聊天应用程序都单独运行的沙箱虚拟智能电话生态系统。 我们对我们建立的系统的业绩和可用性评价强调了在确保现代智能手机不受零点击袭击方面的几个缺陷和基本挑战。 我们在这个经验报告中讨论了我们学到的教训, 并分享关于缺少的部件的洞察力, 以便实现对现代移动设备零点击攻击的无线的安全。