Aging in Place with Google and Amazon

AUTHOR(S) & CREDENTIALS: Jessica Percy Campbell

AFFILIATED INSTITUTION(S): AGE-WELL NCE and University of Victoria, Department of Political Science

ACKNOWLEDGEMENTS: This PhD research has been funded by the Big Data Surveillance Project as well as the AGE-WELL Michael F. Harcourt Policy fellowship. Supervised by Dr. Colin Bennett

What inspired you to begin working on this project?

When my grandfather passed away a couple of years ago, family and friends began suggesting that my grandmother turn to technology to help her live at home alone (e.g. WIFI-enabled security cameras, fall detector sensors, etc). As a surveillance and privacy scholar, that led me down the path of researching digital technology for older adults. I soon discovered that many blogs and websites geared towards older adults and their caretakers were promoting Google and Amazon smart speakers as tools for aging in place. The collective claim is that they can be used as mobility aids, digital companions, medication/appointment reminders, and more. However, the content of interactions with voice assistants can be very personal in nature as they may pertain to health and wellness. Moreover, voice data in and of itself can be considered sensitive, as personal information (such as health status) can be inferred through intonation and other audible qualities. As smart speakers are embedded with voice-activated assistants such as Alexa and Google Assistant, what types of behavioral data are collected for commercial purposes? Moreover, how is meaningful consent to collect older adults’ personal information obtained?

What is your project about?

Put simply, my work questions the privacy and surveillance applications of Google and Amazon voice assistant devices for older adults. As the Internet of Things (IoT) and voice-assistive technology become increasingly pervasive in everyday life, I ask, what are the limits of sensitive data collection for commercial purposes? Google is the reigning champion of “surveillance capitalism”, the practice of turning behavioral data into profit for targeted advertisements (eg. see Zuboff, 2019). Relatedly, Amazon’s behavioral data collection is funneled into its online shopping platform to predict and suggest relevant purchases to users. If older adults are using smart speakers as tools for aging in place through features like medication reminders, home control, and digital companionship, should their data also be collected for commercial purposes? Moreover, in the instance where someone other than the primary user sets up a smart speaker for an older relative instance and taps “I agree” to the terms of service, how is the meaningful consent required to collect behavioral data ever actually obtained?

These questions are to be explored through qualitative interviews with privacy and surveillance experts, focus groups with older smart speaker users in British Columbia, and documentary analyses of the relevant privacy policies and marketing materials.

What are some of the current or future real-world applications?

Raising awareness about the commercial surveillance practices of these companies could be beneficial not only to older adults, but to smart speaker users of all ages. Ideally, this project could encourage users to keep their privacy settings limited, to frequently delete voice recordings, and to set strong unique passwords with 2-factor authentication on all IoT devices. It could also further encourage IoT companies to embrace Privacy by Design (PbD) principles in their products to ensure user privacy is adequately respected.

What are some important policy implications your project may have?

This project also has implications for privacy policy in Canada. In many jurisdictions, companies already face legal limits on collecting behavioral data on children under the age of 13, but adults over 80 have no such extra protections. If older adults are to be encouraged to use Internet health apps and IoT devices for aging in place as an alternative to living in care homes, their privacy should be adequately respected. Currently, notice and consent models of privacy protection are not strong enough to protect individual users from oversharing personal information with companies who then use it for commercial purposes. Ideally, a stronger set of policies at federal and provincial levels could address this issue in a way that encourages more equitable privacy considerations for older adults.