AI-MAPS November 2023 Workshop

Blogpost for the AI-MAPS project by Nanou van Iersel
Helix engagement with surveillance AI

Towards the end of November, I hosted a workshop on behalf of AI-MAPS. Together with my PhD supervisors, I created a workshop around cameras and video material. This workshop began with the idea of cameras as multipurpose technologies. A camera is installed for a specific reason, or on a certain legal basis, but the footage will likely turn out useful for additional purposes. Should it be possible then to repurpose the video material? And is it desirable?

The law offers some guidance on the conditions under which camera material can be used beyond the original purpose. Yet, legal frameworks cannot account for the specifics of every context. So, individuals negotiate legal frameworks on a case-by-case basis, apply it to the specific context, and weigh the different interests at stake. This context, in all its nuances, formed the starting point for this workshop. In other words, with this workshop, AI-MAPS (and I) sought to understand the context in which shifts in purpose occur when it comes to cameras and video material. 

For this, we invited around twenty participants with expertise in cameras through their work. The participants involved represented different organizations, including law enforcement, government, industry, and knowledge institutions. Each guest brought their own type of expert knowledge: technical knowledge of cameras and artificial intelligence, professional knowledge from using cameras for public safety purposes, and/or scientific knowledge of technology impact assessment.

Helix engagement with surveillance AI

The first part of this workshop was dedicated to a set of questions, aimed at understanding the conditions under which the repurposing of camera material occurs. What role do legal frameworks play? And proportionality considerations or non-legal factors like limited time and resources or societal pressure?’. The participants were divided into four break-out groups. Some time (and coffee and cake) later, we reconvened to discuss the questions in plenary. Many insightful comments, questions, and examples from practice have been discussed. To mention but a few:

  • For one, cameras as multipurpose technologies proved to be an accurate description. Shifts in purpose for both cameras and their footage turned out to be common practice. We could challenge this by asking ‘When we can properly speak of a shift in purpose, when the original and additional purposes can be captured under the umbrella of ‘public safety’? At the same time, we want to avoid that anything goes because public safety is too broadly defined. 
  • Second, when one camera can serve multiple objectives, we can start thinking about shared camera infrastructures for different organizations (including law enforcement, firefighters, municipality, and private parties). If the footage can serve multiple organizations, then why not share the camera to begin with? While legally possible, this doesn’t always happen. What are the reasons for (not) doing so when there is no legal inhibition? 
  • Third, law gives a framework, but it still on individuals to interpret and apply that framework in a specific context. This is why ethics is so important; it gives substance to generic guidelines in the law. How do we then nurture personal and professional ethics to make sure that we are making the right choices? For public servants, this question is answered by the oath they took to act in good faith. Yet, there are many more parties with access to cameras – how to ensure we all act in good faith when it comes to cameras and the repurposing of video material? 

These are but some of the questions that stirred our discussion after the break-out sessions – which only become more complex when adding AI to the equation. The second part of this workshop was devoted to a fictive AI-case involving facial recognition software. For this, we set up a panel by inviting four participants to each defend one of the ELSA-perspectives: technical, ethical, legal, or social aspects. By posing (critical) questions to the panel, we collectively anticipated and assessed the impact of AI on current camera infrastructures. A little bit of imagination can tell us a lot about the future!

Altogether, the workshop yielded a rich variety of insights. Following AI-MAPS’ ideology, we learned from practice. I can safely say that my reflection on cameras as multipurpose technologies has only just begun. Once again, I would like to thank all participants for their time and their time and effort.

Related content
Blogpost for the AI-MAPS project by Evert Stamhuis
Evert Stamhuis at the Winelands Conference 2023, Stellenbosch, South Africa
Blogpost for the AI-MAPS project by ginger coons
Left to right: Friso van Houdt, Gabriele Jacobs, ginger coons
Related links
Overview blogposts | AI Maps

Compare @count study programme

  • @title

    • Duration: @duration
Compare study programmes