Alexa, is the skill always safe? Uncover Lenient Skill Vetting Process and Protect User Privacy at Run Time
Alexa, is the skill always safe? Uncover Lenient Skill Vetting Process and Protect User Privacy at Run Time
复制标题
DOI：
10.1145/3639475.3640102
复制
发表时间：
2024-04
期刊：
影响因子：
--
通讯作者：
Tu Le;Dongfang Zhao;Zihao Wang;Xiaofeng Wang;Yuan Tian
中科院分区：
文献类型：
--
作者：
Tu Le;Dongfang Zhao;Zihao Wang;Xiaofeng Wang;Yuan Tian
语音个人助理（VPA）平台（例如亚马逊Alexa）允许开发者在第三方服务器上部署他们的语音应用程序。然而，这种策略给VPA客户带来了意想不到的隐私风险。恶意开发者可以动态改变他们的应用程序的行为，以绕过平台的审查过程。本文旨在系统分析Alexa的语音应用生态系统（即Alexa技能），重点关注行为操纵（也称为技能行为改变）。我们识别恶意技能发布的根本原因，并提出有效保护用户的防御解决方案。首先，我们揭示了亚马逊的技能审查策略以及与他们的审查相关的隐私问题。我们发现，除了技能发布前的技能认证流程外，亚马逊还在技能发布后部署了技能监控方案。我们进一步发现了这种监测方案的局限性，这在以前的研究中没有被探索过。最后，为了解决这些问题，我们提出了一种运行时技能监控方法来检查用户与技能交互时技能行为的一致性。我们的发现建议采取行动，改进VPA技能的审查过程，而不给技能开发人员带来负担，并帮助开发人员遵守政策。•安全和隐私→特定领域的安全和隐私架构；隐私保护;安全和隐私方面的可用性；•以人为本的计算→个人数字助理。随着亚马逊Alexa等语音控制设备提供自动化和便利性，家庭变得越来越智能。通过添加可由第三方托管的语音应用程序（也称为“技能”），设备的功能不断扩展。然而，我们的研究揭示了这项技术的阴暗面——可能会侵犯隐私。恶意开发者可以在绕过亚马逊的审查后修改他们的技能行为，这可能会侵犯用户隐私。尽管亚马逊有适当的预防措施（例如，技能认证和重复监控），但对手仍然可以从漏洞中溜走。在本文中，我们通过现实世界的例子全面揭示漏洞，以便更好地理解，并提出解决方案，有效地帮助用户监控动态技能行为。
Voice personal assistant (VPA) platforms (e.g., Amazon Alexa) allow developers to deploy their voice apps on third-party servers. However, this strategy introduces unexpected privacy risks to VPA customers. Malicious developers can dynamically change their app’s behaviors to circumvent the platform’s vetting process. This paper aims to systematically analyze Alexa’s voice app ecosystem (i.e., Alexa skills), focusing on behavior manipulation (also referred to as skill behavior change). We identify the root causes of malicious skills getting published and propose a defense solution to effectively protect users. First, we uncover Amazon’s skill vetting strategy and the privacy issues relevant to their vetting. We reveal that, in addition to the skill certification process before a skill gets published, Amazon also deploys a skill monitoring scheme after the skill is published. We further discover limitations of this monitoring scheme that have not been explored in previous research. Lastly, to address these issues, we propose a run-time skill monitoring approach to check the consistency of the skill behaviors when users interact with skills. Our findings suggest a call for action to improve the vetting process for VPA skills without placing a burden on skill developers and help developers adhere to policies.CCS CONCEPTS• Security and privacy → Domain-specific security and privacy architectures; Privacy protections; Usability in security and privacy; • Human-centered computing → Personal digital assistants.LAY ABSTRACTOur homes are getting smarter with voice-controlled devices like Amazon Alexa offering automation and convenience. By adding voice applications (also called "skills"), which could be hosted by third parties, the abilities of the devices keep expanding. However, our research reveals the dark side of this technology - possible privacy breaches. Malicious developers can modify the behaviors of their skills after bypassing Amazon’s scrutiny, potentially violating user privacy. Although Amazon has preventive measures in place (e.g., skill certification and repeated monitoring), adversaries can still slip through the cracks. In this paper, we comprehensively uncover the loopholes with real-world examples for better understanding and propose a solution to effectively help users monitor dynamic skill behaviors.