The first part of this series of articles laid out the difficulties of securing devices in the Internet of Things (particularly those used in the human body). Accepting that usability and security have to be traded off against one another sometimes, let’s look at how to make decisions most widely acceptable to the public.
The recent FTC paper on the Internet of Things demonstrates that they have developed a firm understanding of the problems in security and privacy. For this paper, they engaged top experts who had seen what happens when technology gets integrated into daily life, and they covered all the issues I know of. As devices grow in sophistication and spread to a wider population, the kinds of discussion the FTC held should be extended to the general public.
For instance, suppose a manufacturer planning a new way of tracking people–or a new use for their data–convened some forums in advance, calling on potential users of the device to discuss the benefits and risks. Collectively, the people most affected by the policies chosen by the manufacturer would determine which trade-offs to adopt.
Can ordinary people off the street develop enough concerned with their safety to put in the time necessary to grasp the trade-offs? We should try asking them–we may be pleasantly surprised. Here are some of the issues they need to consider.
What can malicious viewers determine from data? We all may feel nervous about our employer learning that we went to a drug treatment program, but how much might the employer learn just by knowing we went to a psychotherapist? We now know that many innocuous bits of data can be combined to show a pattern that exposes something we wished to keep secret.
How guarded do people feel about their data? This depends largely on the answer to the previous question–it’s not so much the individual statistics reported, but the patterns that can emerge.
What data does the device need to collect to fulfill its function? If the manufacturer, clinician, or other data collector gathers up more than the minimal amount, how are they planning to use that data, and do we approve of that use? This is an ethical issue faced constantly by health care researchers, because most patients would like their data applied to finding a cure, but both the researchers and the patients have trouble articulating what’s kosher and what isn’t. Even collecting data for marketing purposes isn’t necessarily evil. Some patients may be willing to share data in exchange for special deals.
How often do people want to be notified about the use of their data, or asked for permission? Several researchers are working on ways to let patients express approval for particular types of uses in advance.
How long is data being kept? Most data users, after a certain amount of time, want only aggregate data, which is supposedly anonymized. Are they using well-established techniques for anonymizing the data? (Yes, trustworthy techniques exist. Check out a book I edited for my employer, Anonymizing Health Data.)
I believe that manufacturers can find a cross-section of users to form discussion groups about the devices they use, and that these users can come to grips with the issues presented here. But even an engaged, educated public is not a perfect solution. For instance, a privacy-risking choice that’s OK for 95% of users may turn out harmful to the other 5%. Still, education for everyone–a goal expressed by the FTC as well–will undoubtedly help us all make safer choices.