Data trusts — a model for Aotearoa (pt I)
Update: see the next in this series over on the Open Data Charter’s blog: Data trusts and trusting openness.
Data stewardship’s a big deal.
As we generate more data, which is more exploited by more companies and governments to further interests which definitely aren’t ours, the idea that data needs trusted stewardship to generate real benefit is all the rage. Genuinely and for trustwashing purposes.
_How_ to do this, however, is far from easy. And a recent concept may have at least some of the answers, some of the time: data trusts.
What on earth is a data trust?
The concept’s still a very new one, so there’s no one definitive definition.
A current working definition I quite like is that, essentially, a data trust is a form of data collaborative, “ a legal structure that provides independent stewardship of data”[1].
There’s also some conversation about whether data trusts can or should have fiduciary duties towards trust members, but the purpose of this post isn’t to dive into the detail and the debate(s) — as fascinating as they are.
Instead, this post serves to point out some possible use cases for Aotearoa. While many I’ve spoken with have assigned data trusts here to the too-hard basket, I don’t believe that’s the case*.
They — and other forms of data collaboratives like data collectives — could be a tremendously powerful tool just at the time when we need them.
To head off increasing data capture by a few large (and generally exploitative) people and companies**, allow better flow of data to drive true social / environmental / economic benefit for us all, and retain or even improve our ever-waning trust.
What follows is a brief, non-comprehensive list of how they could help New Zealanders.
Data trusts in Aotearoa
I’m most excited about data trusts in terms of two primary mechanisms:
- as a means for individuals and communities to have more say / control over how our data is used, and by whom
- as a means for private companies to share data through a trusted third party intermediary, where they can’t / won’t directly due to competitive concerns.
While I can see ways for pretty much every sector in NZ to potentially benefit, here are some primary use-cases. Note that I focus strongly on ethical ones.
Māori data sovereignty [3]: groups and organisations — including government — throughout NZ either already see data as taonga, or are beginning to. Data trusts could be a valuable way to better unlock data’s value for New Zealand, while still protecting it.
Medical data (as opposed to 23andMe’s approach**): kinda does what it says on the box here. Rather than the current all-or-nothing situation, we could decide who could do with which parts of data, with potentially _huge_ positive benefits for medicine.
“AI”: NZ’s artificial intelligence (AI***) industry isn’t doing as well as it could and should be, and one of the reasons for this is lack of access to data. Data trusts could potentially be a way for those with data, to share said data with AI-using organisations both more easily and more ethically.
Smart cities and digital twins: smart cities and their next-gen digital cousins, digital twins, both require and generate ever-growing amounts of data. One of the many major issues around this is how to _share_ the data so that it benefits rather than exploits people and systems. Data trusts, as with our Crown Research Institutes (CRIs), would be a way for government, community, NGO and private sector organisations to steward and share data. Civic data trusts in particular could be really useful here — I know Adam Beck [4] is doing some lovely thinking on the subject, for example.
Digital identity: as the debate hots up over who should control our digital identities, and how, data trusts could provide an alternative to some of the current, suboptimal(to my mind) models and possible options, by giving individuals control over the components of their digital identity for different uses, without requiring them to try manage and control every aspect, every time (a nigh-on impossible task even for the digitally highly literate). Certainly, it could mean the Googles and Facebooks (and problematic governments) of the world could stop being these arbiters. #selfsovereignty
Crown Research Institutes: New Zealand’s CRIs hold a veritable treasure trove of data, but commercial agreements and intellectual property sensitives — as well as technical debt — make it difficult for them to share data. Data trusts could serve as trusted third party intermediaries to allow equitable data exchange.
Government: data trusts could provide a means for us to to give consent to government to share data about us — for different use cases — amongst themselves, to build better services and delivery.
Agriculture and other primary industries (and more generally): water quality, environmental pollution and animal ethics are huge issues in and around agriculture — one of our primary industries. Data trusts could help our organisations share and use data better on wider levels, for example catchment sustainability, in such a way that they’re not concerned things will come back to bite them in the arse later. Fear of how things might be perceived is a _major_ blocker here to organisations sharing data, and thus to sectors improving effectiveness, efficiency, behaviour and thus reputation.
Conservation: a key issue with sharing conservation-related data is that it could unintentionally _harm_ affected species, for example by showing hunters and poachers where target populations might be. Data trusts could help to improve data sharing while still protecting the rights of other species. [Shoutout to Victor and wildlife.ai at this point!]
Colour me intrigued — what next?
As I mentioned, this is a brief post.
I’ve not started waving my hands around about data trusts and the sustainable development goals, the concept of data ownership (and why it’s so problematic), who should and should not be running these, and so forth. There’s a bunch of reading material at the end of this post.
Nor have I gone into how one might build the systems for people to find these things, for the data sharing to actually occur, and so forth (I have Thoughts and Ideas there too).
My collaborator Tim Packer and I are currently collecting use cases for a pilot trial or two. Specifically, we’re looking for:
- datasets to which people might like access, or
- datasets which people have but are having trouble sharing, or
- specific problems people want to solve.
And yes, we’re definitely taking lessons from the pilots the Open Data Institute have run so far.
If any of this sounds interesting, give me a shout. Let’s get this thing going.
— — —
* Indeed, Data Ventures, for whom I used to work, counts as a form of data trust.
** Don’t get me started on Sidewalk Toronto and their recent use of the “data trusts”. I’m going to be watching veeeery warily before I decide to trust them. I’d argue they’re…not a good example to be following.
*** “It’s not ****ing AI” is one of my more common rants, but for the sake of simplicity, let’s just use the common term here.
— — —
References and further reading
[1] https://theodi.org/article/odi-data-trusts-report/
[2] http://freakonomics.com/podcast/23andme/. To say her and her company’s attitude and approach had me spitting mad would be an understatement.
[3] https://www.temanararaunga.maori.nz/ng-mahi
[4] https://www.linkedin.com/in/adampbeck/
A bunch of links over at diigo.com/user/aimee_w, under the tag #datatrust.
For more on open data / smart cities / civic tech etc:
