With the rise in discussion about and implementation of AI in health and care, now is a good time to start a conversation about governance. AI is going to be a big deal and has big consequences. The sooner we get the governance conversation going, the better off we will be as clinical leaders in informatics.
Here’s a link to an article to get the conversation going. Over to you now.
Hi Karen, I absolutely agree. Some of the thinking in this space is frighteningly naive. Maybe we can add this to the list of discussion topics for CLiN but I would be keen to be involved regardless?
Having thought about these issues a lot recently - it is probably appropriate for this to concurrently discussed with data governance.
Is there a group thinking about this in the clinical space.?
Matthew
The UK Faculty Of Clinical Informatics has an AI subgroup which is having its inaugural meeting in London on 4th December 2019, obvs not that useful for NZ people as they won’t be able to attend, but hopefully there will be some shareable outputs.
I’ve been thinking a bit about how one might regulate and accredit a workforce of AIs, comparing it somewhat to how we regulate and accredit human intelligences, with all their variations and motivations. There are clearly significant differences between the two, and not always clear what the (technical/cultural/sociological) solution is…
I’m hoping to have my talk on this accepted for Digital Health Rewired 2020. If it is accepted then i’ll crack on at actually writing it ![]()
How do you see governance working here? Given existing data and information governance is so delicate already, and no existing framework in the country ( canterbury and Auckland inclusive) don’t have any major limitations, the aspect of AI causes a significant amount of issues around ethics, social license, social contract for responsibilities, data sovereignty, liabilities when things fail, and not to mention the privacy aspects.
As part of my Precision Driven Health project, navigating all aspects of this in the NZ primary context have thrown up a LOT of problems, misconceptions, limitations and unforeseen ethical considerations that means AI, even in the most rudimentary sense, has significant legal/regulatory challenges that is one step higher than the current understanding of the data governance issues already in NZ.
Agree this is such a key element to occurs across many threads. In the NZ context, what is the thinking around data governance and sovereignty? With the te Tiriti o Waitangi in mind, it seems we must consider sovereignty if we are going to talk about governance. I agree that if we don’t have plan for this, there are huge risks, particularly of worsening inequities, moving forward.
Great conversation, thanks for sharing.
Here is a recent article on this topic:
It emphasises governance but doesn’t mention indigenous sovereignty which I think is critical.
Should there be a single, central governance approach? Or continue with existing regional approaches?
Hi @matthew.strother
Yes there is a NZ Data governance forum on Yammer
@karenblake has been one of the leaders of this.
They have a monthly teleconference on the first Thursday of the month (so next one in a couple of days time)
There’s also a data workshop on the Monday of the HiNZ digital heatlh week (Mon 18 November)
I haven’t kept up with the discussion on this group in the last few months, but perhaps @karenblake can tell us if and how AI governance is being considered as part of this?
There’s a new governance model for the application of AI in healthcare that was published in AMIA which came out of Australia just a few days ago.
It seems like a good foundational model. From my perspective, it doesn’t include education as part of the model which is key to developing capability.
Interested in what other people think.
Hi all, we spoke about this at the NZ health data forum this morning- it’s certainly am emerging issue. The sector is implementing data governance processes/committees across most regions/DHBs- and this remains at low maturity given the newness of data governance as a discrete disciple/construct. I would hope that people can leverage off existing data governance structures, and adding in any governance gaps: legal and consumer were raised today as gaps.
Another option would be to coordinate this from the centre, with AI governance being included in the work Simon Ross is doing with data governance at the Ministry.
Hi @karenblake. I am keen to be involved in the conversations around data governance but apologise for not having been able to attend the meetings.
Interesting article from today on this:
Interesting privacy perspective.
