Page 39 - U4SSC Guiding principles for artificial intelligence in cities
P. 39
8 Conclusion
Cities are either deploying or are planning to deploy AI systems to address their urban challenges. It
is important to establish guiding principles for AI systems to ensure reliability and trust. Some of the
conclusions from the framework and the real-life case studies in this deliverable are stated below:
• There is no “one size fits all” approach for guiding principles for AI in cities;
• It is important for cities to consider implementation issues in order to successfully ensure
compliance with their AI principles;
• It would be beneficial for cities to segregate data and algorithms and individually consider both
for principles compliance;
• Urban challenges and priorities, city administrations’ smart sustainable city strategies, inhabitants’
urban requirements can act as a viable demand for AI principles as part of a broader context
for AI implementations;
• City administrators have a wide range of tools at their disposal to encourage and incentivize
implementation of AI principles defined as enablers in this framework document (e.g.,
regulations, policies, awareness, start-up ecosystem);
• Exchange of knowledge at the local, regional, and international levels will help develop AI
principles formulation and implementation and will also increase its sustainability in the long
run;
• Cities have a wide spectrum of policy alternatives depending on how flexible or obliging they
want to be in their AI principles implementation;
• Three governance options for the extent of centralization have been provided to cities in
regulating, implementing, and assessing their compliance with respect to AI principles;
• Guiding principles for AI is a relatively novel topic and may benefit significantly from capacity-
building and research and development (R&D) programmes; andF
• Cities can capitalize on AI principles by turning it into a viable economic sub-sector, while
simultaneously utilizing them in addressing and solving their own urban challenges.
29