Blog 11: Complete Delete

The life cycle of data

Case Study:
Complete Delete: In Practice, Clicking ‘Delete’ Rarely Deletes. Should it?

This case study talks about the life of data and the uncertainty behind it. Data is often deleted in hopes of it disappearing without a trace, yet most systems have some way of maintaining backups or copies of data, with or without users’ awareness. Different types of policies are discussed and the ways that they can be beneficial or harmful.

How can we inform users that their data might remain after they delete it? First off, I think terminology should be changed or edited. A warning that says “are you sure you want to delete this?” should be used if something is truly being permanently deleted from all sources without any further effort. If we instead say “do you want to remove this local copy”, users would be better informed about what is actually happening to their data. If I have a sensitive picture that I don’t want to remain on my phone, I would want to know that what the system is telling me it’s doing is accurate. For older data that might be hidden, I think it would be beneficial to have a monthly notification for files that haven’t been viewed or interacted with at all in the past month. This could be changed to every 3 months or so in order to not pester users as much. I have seen systems that notify users of games they haven’t played in a while, which reminds them of the data they are currently using.

Deleting information isn’t necessarily always the best case scenario though. How do we balance the idea that information should be accessible and unaltered for a long time with the idea that some data should not be kept? I think it matters what the data is and how it was created. If the data is a photo a user takes and wants to delete, it doesn’t have much purpose having information permanence. It might be nice to have some backup in case they didn’t mean to delete it or regret it, but some systems have a secret 30 day trash bin before it’s fully deleted. In contrast, certain things like records of history and criminal records should be harder to delete since they could be altered. Having multiple copies would allow for double checking, and also prevent editing of its contents.

So how could we design a system that addresses these issues? One idea is prompting users during setup with questions regarding their data. Some examples would be whether or not they want to save space by uploading things to the cloud, as well as how long they want data to be kept after “deleting” it. Another idea is to inform users every time they attempt to delete data about the status of the data. This could be something along the lines of a message saying “This file is still available in the _ directory” after removing it from the main directory.

Would a system using cryptographic erasure allow for images to be on websites and be secure? No, there would be many measures people can take to get their own copy of it. As stated in the article, people can always take pictures of a cell-phone screen, allowing for distribution that doesn’t require decryption. These websites could potentially watermark images or edit them somehow to make them different from the original, but this could defeat some of the purpose of the image. The original will still be able to be deleted by deleting the key, but its edited versions could easily stay on the internet forever.

After reading this case study, I wonder if there should be policies set in place to guarentee erasure of certain data. What would the extent of this policy be and would there be exceptions?

For some people, certain things could damage their image in society, or possibly cause harm to other people. Would it be beneficial to have legal policies made to protect people’s right to data erasure or would it complicate law enforcement’s ability to protect society? This could potentially cause more awareness on data privacy, as well as improve user’s trust in technology. I think that if there were policies set in place that there would need to be very clear indication of when data will be stored and impossible to delete permanently.

This case study brings up important decisions that users and developers must make for data. If we don’t make changes to the current structure of technology, we are at risk of privacy concerns. We have already seen many cases of data breaches and people having information or photos exposed. Companies need to be more transparent about the state of data, in order for users to know of its existence and their options. Data is something that many people nowadays don’t think too much about, so we must also educate people about how their data is stored and where it goes.