SkillVet: Automated Traceability Analysis of Amazon Alexa Skills
- 18 November 2021
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 20 (1) , 161-175
Abstract
Skills are essential components of Smart Personal Assistants (SPA). Their numbers have grown rapidly, dominated by a changing environment that has no clear business model. Skills can access personal information and this may pose a risk to users. However, there is little information about how this ecosystem works, let alone the tools to facilitate its study. In this paper, we present a systematic measurement of the Alexa skill ecosystem. We study developers practices, including how they collect and justify the need for sensitive information, by designing a methodology to identify over-privileged skills with broken privacy policies. We collect 199,295 skills and uncover that around 43% of the skills (and 50% of developers) that request permissions follow bad privacy practices, including (partially) broken data permissions traceability. To perform this kind of analysis at scale, we present SkillVet that leverages machine learning and natural language processing techniques, and generates high-accuracy prediction sets. We report several concerning practices, including how developers can bypass Alexas permission system through account linking and conversational skills, and offer recommendations on how to improve transparency, privacy and security. Resulting from the responsible disclosure we did, 13% of the reported issues no longer pose a threat at submission time.Keywords
All Related Versions
Funding Information
- Engineering and Physical Sciences Research Council (EP/T026723/1)
- Ramón y Cajal (RYC-2020-029401-I)
This publication has 32 references indexed in Scilit:
- Continuous Authentication for Voice Assistants2017
- Systematization of Knowledge (SoK): A Systematic Review of Software-Based Web Phishing DetectionIEEE Communications Surveys & Tutorials, 2017
- A Privacy Assessment of Social Media Aggregators2017
- Designing Effective Privacy Notices and ControlsIEEE Internet Computing, 2017
- Reevaluating Android Permission Gaps with Static and Dynamic Analysis2015
- Social Networking Privacy: Understanding the Disconnect from Policy to ControlsComputer, 2012
- Automatically securing permission-based software by reducing the attack surface: an application to Android2012
- An extensive experimental comparison of methods for multi-label learningPattern Recognition, 2012
- Detecting spammers on social networks2010
- Learning to detect phishing emails2007