AI as Public Service: Accountability & AI-Registers
Last month Helsinki and Amsterdam announced the launch of their open AI registers. This is the first time such a service has been offered. The goal of the open AI registers is to ensure and demonstrate responsibility, transparency and security concerning urban AI data solutions. As of now, the selection of applications in the register is limited. However, they are expected to expand and include all the cities’ algorithms in the future. The registers include an overview of the applications, including the data they use, their operating logic and how the applications are governed. Click here for the Helsinki register and here for Amsterdams’s register.
The open AI registers demonstrate a way to deal with the complex urban environment in a way that uses AI as a means to an end. This helps relieve pressure from the system, since less human interaction is needed to coordinate processes. To ensure that the AI algorithms are used in an ethical fashion, the implementation includes treating the underlying system as open source. This ensures that it complies with the open transparent and secure policies that Helsinki and Amsterdam have for other processes in the municipalities. An example of one of the projects found in the AI register is the report system for litter. Before, citizens that saw an instance of litter would make a report and assign the litter to a list of categories. However, due to the number of categories, litter was often categorized wrongly, leading to inefficiencies. Now, with AI, it is possible to use an algorithm that classifies the litter in question.
The open approach also serves to build trust with the citizens, since they can see how the data collected is processed and how the code is built. Additionally, the very fact that the code is public and can potentially be seen by anyone, ensures that developers treat the system with scrutiny. It ensures that they consider the ethics surrounding technology when they implement changes. The article here outlines the potential for both AI registers to establish trust and and transparency. There are benefits to beginning this project on a city scale as opposed to a national scale. Municipalities have the flexibility to tailor the approach to the city. There are also challenges Most of these challenges are issues that have a more global scale and thus require a more global approach. However, through consistent cooperation between cities, these can be overcome.
By implementing similar systems in their own cities, municipalities can join a growing network of institutes that incorporate AI technologies to increase the efficiency and effectiveness of their processes. The more cities adopt this technology, the more integrated it ultimately becomes. Showing that technology serves to help our processes and that it adds value in a transparent, open and secure way can show why AI is important. Building a strong foundation for incorporating AI now will lead to easier implementation of technological advances down the road. The municipalities are welcoming suggestions or improvements on the beta version of the register.
Written by Sanjay Ghosh