Trust and Transparency in the Age of AI and Algorithms

1. Ethical Framework for AI in Cities
Purpose: The webinar introduced an ethics and inclusivity framework developed by Demos Helsinki, designed to guide cities in piloting AI solutions, especially for marginalised or vulnerable groups.
Core Principles:
Co-creation: Solutions should be developed with and by marginalised communities, not just for them.
Accountability: Cities must be prepared to take responsibility for both the successes and failures of AI pilots, ensuring no vulnerable group is left behind or blamed.
Resource Allocation: Adequate resources and support ecosystems are essential for meaningful inclusion and ethical piloting.
Reflection & Checklists: Continuous reflection and practical checklists help cities and funders ensure ethical standards are met throughout the AI lifecycle.

2. Case Study: Mijke, the Social AI Assistant (Eindhoven)
Objective: Mijke is a chatbot designed to humanise digital systems and scale human connection, especially for people lacking digital or basic skills.
Key Features:
Judgment-free, accessible guidance via familiar channels (e.g., WhatsApp).
Multilingual, step-by-step support to navigate complex local services (health, finance, work).
Co-created with marginalised groups and experts, ensuring the solution addresses real needs and builds trust.
Lessons Learned:
Society-driven approach: Start with lived experiences and local challenges, not technology.
Trust by design: Transparency, privacy, and accountability are critical; involve users at every stage.
Heterogeneous needs: Recognise that marginalised groups are diverse; one-size-fits-all solutions rarely work.
Impact over adoption: Focus on validated human experience and real-world impact, not just user numbers.

3. Challenges & Tensions in AI Adoption
Municipal Officers:
Fear of job loss or role change due to AI automation.
Need for upskilling and involvement in AI design to ensure compliance and accountability.
Solution: Municipalities should co-create with tech providers and invest in sustained knowledge transfer.
Policy & Governance:
Current AI policies often lag behind technological evolution.
Need for modular governance structures to address different AI instances (e.g., generative AI, embedded AI).
Joint procurement and regional cooperation can help smaller cities pool resources and expertise.
Civil Society & Inclusion:
AI must address real problems and close gaps (e.g., digital divide, access to services).
Citizen engagement is crucial for trust and acceptance, especially regarding data use and privacy.

4. Actionable Takeaways
For Cities:
Adopt and adapt ethical frameworks for AI pilots.
Invest in co-creation and capacity building for staff and citizens.
Prioritise transparency, accountability, and inclusivity in all AI projects.
For Tech Providers:
Involve municipalities and end-users early and often.
Design for accessibility and real-world impact, not just technical efficiency.
For Funders & Policymakers:
Allocate resources for inclusive, long-term AI projects.
Support knowledge sharing and collaboration across regions.

5. Closing Thoughts
AI is a tool, not a solution in itself. Its value lies in how it is designed, deployed, and governed.
Trust and transparency are foundational for public acceptance and successful AI adoption.
Collaboration across sectors (municipalities, tech providers, civil society, funders) is essential to ensure AI serves all citizens equitably.