Data unions

Much ado has been written about giving users more control of and value from their data. Without spending too much time on the forces beyond this movement, people have lost trust in the institutions stewarding their data and feel a sense of inequity as these institutions benefit off what’s perceived to be an extension of individuals.

I like the idea of data unions. Just like credit unions pool money together for the collective benefit of their members, data unions would pool data together. These would be owned and managed by their members and any profits would go back to their members as well. By pooling data in a model like this you could achieve the scale necessary to generate meaningful results with AI, but still members would retain autonomy and be able to profit from their data.

A place where this might have a natural fit is with patient advocacy groups. These are organizations that exist to combat a particular disease, disability, or a group of diseases and disabilities. In some communities patient advocacy groups are critical; they serve as a forum to connect with others, a hub for information, and an advocate for their constituents. A few groups are very well funded and are directly engaging in research via venture philanthropy already.

What if they extended their advocacy to deciding how and when their constituents’ data should be used? They already act as gatekeepers to their members, choosing which organizations get access and which clinical trail recruitment calls get passed on. It seems a short leap forward to have them vetting proposals on how to use their constituents’ data on their behalf. Particularly for more esoteric diseases there is a burgeoning need for data, and those affected want to contribute but don’t know how or simply can’t. Advocacy groups act would act as a natural centralizing body for a group of people that is otherwise disconnected and disparate.

I’ll readily admit I don’t know what the governance structure for this would look like. Credit unions, to my knowledge, mostly operate by a committee elected by their members. Something similar could work here. Forward thinking data unions could adopt quadratic voting. As with any organization it becomes more and more difficult to represent your constituents as they scale. A data union that could negotiate with Facebook is going to be orders of magnitude larger than a rare disease advocacy group, and will probably do worse at representing everybody under its umbrella as a result.

As a last note blockchain technology could be the infrastructure that enables this. It’d provide the technology to let people both take control of their data and enter into a contract with other parties for them to manage it on your behalf. Whatever governance structure these unions choose could be entrenched with smart contracts and on chain governance. Payments could flow automatically with smart contracts and data usage could be recorded and audited on an immutable ledger.

If anyone wants to work on this then please get in contact with me, I’d love to work with you and I’ll be the first person to sign up.

Connect with me

You can also follow me on here on Medium or on Twitter. I appreciate feedback!

Data unions was originally published in Hacker Noon on Medium, where people are continuing the conversation by highlighting and responding to this story.