Amsterdam and Helsinki became the first two cities in the world to launch AI-based registers that log algorithms used in municipalities. Finnish developer Saidot created the registers used by both cities.
The cities announced this development at the New Generation Internet Policy Summit organized by the European Commission.
According to the Government AI Readiness Index 2020, Netherlands and Finland are the most prepared to adapt AI into government services.
Currently, the AI registers in the two cities contain only a handful of applications. Helsinki allowed its AI to have a chatbot service, along with a management system for the city’s library.
On the other hand, Amsterdam used the AI system to control the city’s parking protocol. Specifically, the technology will analyse issues of public space.
Both cities plan to include algorithms for all public services in an AI register. Most importantly, these systems are indicative of how AI will enter and affect our lives.
Issues of transparency
The AI registers in Helsinki and Amsterdam fall under the bracket of ‘automated decision systems’.
This is not the first attempt at automated decision systems. Notably, the city of New York tried creating a task force to assess the creation of automated decision systems. Unfortunately, the project was dropped in 2019 after the task force stated problems of transparency and of lack of access to the city’s algorithms for governance.
A joint study by Stanford-NYU on the use of AI systems by governments iterates that poor performance, or algorithmic bias can lead to loss of trust between governments and citizens.
For City municipalities to succeed, they will have to facilitate a completely transparent transfer of governance algorithms to AI registers.