AI Ethics in Practice: Designing for Citizens

Henriette Friis

Invited lecture by: Meeri Haataja, CEO and Co-Founder, Saidot.ai, March 11, 2021.

Last week we had the pleasure of hearing from Meeri Haataja who talked to us about public AI registers, like the ones seen at the City of Helsinki. In her talk, Meeri highlighted how AI systems need to be continuously improved, and should never be considered fully done.

Meeri talked about Saidot.ai and how the start-up is helping to ask questions such as “How do you ensure equality and avoid harmful biases in your system context?”, “What potential risks and harms may the use of your system cause?”, “How does the system work” and “Who does it impact?”

Meeri also discussed challenges of getting meaningful feedback from citizens and how to implement it in the AI systems. A topic that this week’s student team incorporated in their participatory workshop following the lecture.

In the workshop the participants were also asked to put themselves in the role of some of the key stakeholders, namely citizens, domain experts, public sector organisations and technology providers. This led to some really important discussions about policy, trust, explainability, funding, education, feedback loops, and, of course, power.

Suggested Readings:

  • Drobotowicz, K., Kauppinen, M., Kujala, S., “Trustworthy AI Services in the Public Sector: What Are Citizens Saying about It?” To be published in the Proceedings of the 27th International Working Conference on Requirements Engineering: Foundation for Software Quality (REFSQ), April 12-15, 2021. (PDF)

Additional References:

Leave a comment