MAINTAINING PRIVACY WHILE VOLUNTEERING DURING THE CORONAVIRUS PANDEMIC
The tendency in a crisis situation is to help out, it’s our nature. We all want to contribute to the common good when under threat. But there’s a right way and a wrong way to volunteer. In our rush to support those in need, our neighbors, it is important to safeguard our own privacy and protect the exchange of data between those in need, and those who want to fulfill that need.
Mutual Aid, is one new program that has sprung up recently to provide a platform for pairing volunteers with those in need. It is a variation on an old idea in the grassroots organizing space, and very well intentioned. There are, at last count, over 650 Mutual Aid pods currently operating in the USA and around the globe. Each implemented in a unique manner by the organizations or individuals responsible for initiating the effort.
Regrettably, some pods have not been thoughtfully designed nor elegantly executed. Inexperienced volunteers (“community organizers”) are starting these local Mutual Aid chapters around the world and some, perhaps many, are not well informed about privacy and how to protect personal data during emergencies such as this pandemic. The result is that the already widespread effort is, in some cases, exposing volunteers and supplicants personal information to the entire online world. And that’s a problem.
Looking to provide their membership with avenues of service and support, some organizations like AARP along with its partner United Health Care promoted the Mutual Aid platform to a pilot program including some members, the author included, in emails and on their website. This, after vetting most of the pods and conducting a thorough legal review. But sometimes even the best efforts miss a potential problem area, and that’s what happened with Mutual Aid, as AARP’s vetting process, as good as it was, missed a few Mutual Aid pods that were (and still are in some cases) using open, public spreadsheets to exchange personal information. One such pod exists in the San Francisco Bay Area, and another outside of Sacramento, California.
The problem is that Seniors are particularly subject to fraudulent exploitation, and often are the victims of crime — cybercrime included. So how does a prestigious, national organization like AARP and its health insurance partner United Health Care respond when advised that their members may be sharing their personal information in an online, public spreadsheet?
The author is pleased to report that once apprised of the potential problem, both AARP and United Health Care took immediate steps to mitigate the risk. Jason Young, AARP’s Senior VP of external relations explains: “Our Community Connections platform launched 22 days ago as a beta project, and we continue to improve it,” said Young. “It’s a platform for people who need help, or who want to offer it, and thousands of people are already being helped. We never intended to link out to open spreadsheets, and we have corrected that.”
To be specific, here are the three corrections that Young outlined in an email to the author:
- We have delisted all mutual aid groups that use spreadsheets for their intake process.
- We have boosted the warnings and content on the site to help users be aware and protect themselves from possible fraud.
- We have added an interstitial so that users understand and acknowledge when they are leaving the site to go to third-party mutual aid groups.
From this author’s point-of-view, the corrections offered by AARP are responsive and entirely in-order, given the circumstances. Their rapid response to the issue is to be commended.
It is important to understand that Mutual Aid is an idea, not a centrally administered program. The idea is to connect individuals who have a specific need with those who can help fill those specific needs. Again, a worthy and well-intentioned goal. But some individual implementations of the idea invite abuse in this author’s opinion. And there are much better alternatives available. Like the grassroots volunteer developed “Adopt a Neighbor” approach that is currently deployed and being successfully used in Ashland, Oregon, the site of Southern Oregon University and the Oregon Shakespeare Festival.
It is accurate to observe that the Mutual Aid system architecture is not laser-focused on geographic proximity like Ashland’s Adopt a Neighbor. By focusing on geographic proximity, and limiting volunteers to just three neighbors, the Ashland, Oregon program seeks to protect privacy and limit data exposure. Not true with Mutual Aid.
One of the weaknesses with Mutual Aid online community is these well-intentioned volunteers are not getting much concrete guidance on data privacy and reducing the potential for accidental spread and misuse of personal data and the potential for fraud. The tools the system admins are cobbling together to get this done in a hurry don’t make privacy and fraud-prevention easy. This is where tech companies and organizations with strong IT departments can and should step-up to provide eager volunteers with proper guidance and the tools necessary to ensure user protection.
Why? Because users of Mutual Aid may risk submitting personal requests for financial help via a Google Form, and the spreadsheet may be open to the whole online world. That’s a potentially big problem, because on the public spreadsheet this author cited, people are openly sharing information such as their sexual orientation, citizenship status, email addresses, and donation amounts, visible to all, and that could make them a target of criminals and fraudsters. Anyone, including a thief or fraudster, could submit a fake request and accept funds from an innocent donor, thereby taking away from someone that actually needs the help. Americans have already lost well over $12 million to virus fraudsters. Silicon Valley, now is the time for you to step up and offer support!
Given the gravity of not having an income, being sick, and feeding their children, the author can see why people are sacrificing their privacy. But they don’t need to open themselves up to exploitation in the process.
Fraud and abuse can and will start happening with something at this enormous scale, and it is clear to the author that some Mutual Aid organizers who are posting some of these spreadsheets online are not technically or privacy savvy. It is important to note that many of the Mutual Aid spreadsheets do have some privacy and data protection protocols in place, now the few that do not need to follow suit.
One of the goals with Ashland’s Adopt a Neighbor program is to grapple with these issues and incorporate more “privacy by design” into the system. To some extent, the authors have already done that with geographic proximity and by limiting the maximum number of neighbors a volunteer can assist to three. That means that volunteers are only getting the private data of three people, and they live in the same community.
With Adopt a Neighbor, the community organizers and the staff of the partnering organization are aware of their responsibility to protect the personal information and not use it for any other purpose.
This is what the authors of the Ashland Adopt a Neighbor program write in their Github readme file:
The principle of limiting access only to the least amount of data needed to achieve our goals can be applied even in an emergency like the Covid-19 epidemic. To minimize the potential for personal data to be abused on the internet, programs should be designed under the same principle we use to minimize the collateral damage in the event of a breach of the software system. We can and should design software solutions that only expose user data to those who are most likely able and/or willing to help, thereby achieving privacy by design. This concept can be extended to “vetting the kindness of strangers” and “trust but verify.” The Adopt a Neighbor should consider how to provide the tools community organizers need while also making it easier for them to put limits on the exposure of personal information. And by design, Adopt a Neighbor only exposes a very minimal amount of information to the average participant because they are only matched, at most, with three other neighbors in their immediate area.
The Adopt a Neighbor program in Ashland, Oregon is now starting to build a more robust and sophisticated system, but they need help to recruit others so this can be more easily replicated and scaled to many other communities and larger cities. Like Mutual Aid, the Adopt a Neighbor program is still more of an idea than a full-blown automated system. The author of the system has made his work open-source and uploaded the details, and a code repository, to GitHub. His team, this author included, is actively looking for technical and organization partners to move to version 2.0. To that end, starting a 501 C-3 nonprofit to manage the effort is being considered. “We would also consider working with an established nonprofit or an existing entity with a robust IT department to take ownership of the project and help us move it forward,” said Mica Cardillo, the software engineer leading the Adopt a Neighbor technical team.
It’s time for a system that values privacy and builds-in data protection and the Adopt a Neighbor program does just that. Neighborhoods, cities and counties must conduct proper and thorough due diligence before committing to pandemic need-pairing and volunteer programs. It’s a crisis, but that doesn’t mean we need to sacrifice our privacy and share our personal information just to help.
