Recommendation 8: Emerging Technology Ethics Advisory Panel
Q 1 What is your Recommendation?
Create an Emerging Technology Ethics group, to be administered by the Wikimedia Foundation, and to consist of representation of paid staff, affiliates and outside experts to review new areas of technology and identify and publish recommendations and guidelines for potential social or ethical concerns raised by the technology. These would be published annually as specific findings of fact and related recommendations to all Wikimedia developers (staff, volunteers and 3rd parties) for how to engage with emerging technologies in a way consistent with the movement’s values, and to identify new technology risks to our users or platform.
Q 2-1 What assumptions are you making about the future context that led you to make this Recommendation?
If Wikimedia is to become a long-lasting core of a free knowledge infrastructure, it will have to assess and incorporate new technology developments. In the time since Wikipedia launched new technologies such as mobile touch-based computing, API service product and machine learning (to name only a few) have grown to become essential parts of the internet. These new developments, and the many more to come, present both opportunities and risks to Wikimedia as a movement and as a technology product. Because our values and ecosystem place unique constraints and open unique applications for these new technologies, the movement as a whole, and particularly the paid staff need a venue and process to discuss and digest new technologies and to make recommendations about their use appropriate to our movement’s needs.
Q 2-2 What is your thinking and logic behind this recommendation?
Currently, discussion and consensus formation for new technologies is ad-hoc and may or may not include experts on technology ethics and societal impacts on technology. Or even where they do, it is often left to the participants to share and advocate for their principles. This is admirable but, if we are to become a long lasting, essential, but values constrained technology movement it won’t scale or be able to keep up with the diversity of new technologies. A recent example of this is in the area of ethical AI. The Foundation’s Scoring platform team and design researchers have many years of thinking about and applying machine learning to editors tasks. They even recently published a draft set of recommendations for ethical AI use in Wikimedia. However, as these are the focused efforts of a small handful of staff and volunteers, their recommendations may not be heard or applied in other areas. By creating an expertise based board, and requiring them to publish regular recommendations in this area, we can elevate these discussions to give them more prominence and governance impact. Additionally this group may examine areas of technology where no paid staff is working, but where a Wikimedian analysis of the risks and opportunities would make those areas more relevant and less problematic for potential staff work or volunteer efforts. An example of this would be block-chain technology. No paid staff of any affiliate are working on block chain technology, which means it is essentially an open field. However block chains carry risk and could be beneficial or harmful to some of our core needs as a technology platform. By having a dedicated group examining and providing guidance, we can fill in some of these open gaps.
Q 3-1 What will change because of the Recommendation?
First, a new movement body, either a sub-group or special purpose group will be created, with administrative support from the Wikimedia Foundation. This new group will be mandated to produce an annual report, identifying and then stating findings and principles in major new areas of technology.
This published global expert consensus will then provide a base understanding that paid staff, groups and individual volunteer developers can use. This will provide a more coherent movement-wide perspective on new technologies, and give moral authority to foreclose or encourage certain technology applications. Although the recommendations would not be binding, it would put contributors violating them in the position of thinking through and justifying why their work remains ethical and within our values.