Cansu Canca was one of the guests at Sunday’s IFA+ Summit sessions, delivering a speech entitled Ethical AI: Everyone Wants It, No One Has It. We asked the AI Ethics Lab Founder and Director about the ideas behind her presentation.

My work is about how to create ethical AI. In AI Ethics Lab, we developed a model for achieving this within institutions, whether they are private companies or public research entities. Our ERD model (Ethics in R&D model) focuses on three key aspects of the AI research, development, design, and deployment processes: Training developers and scientists in ethics, embedding ethics analysis within the product design and development process, and developing institutional principles and processes for recurrent ethical problems.

Can you tell us a little about the role of the AI Ethics Lab?

Everyone wants to talk about ethics and run projects on it but no one seems to realise that this requires expertise in this field called Philosophy. What still distinguishes the Lab is the expertise. It is of utmost importance to me to have the right experts in each and every project. That means the Lab’s work involves philosophers, computer scientists, legal experts, and others.


How do the issues at stake vary across cultures and regions?

Each region and country has its own concerns due to their culture, history, and political situation. However, this does not mean that the core problems differ. Concerns for privacy, individual decision-making, well-being, and fairness hold regardless of your trust in your government or the composition of your society. Their demands and their importance vary depending on the exact circumstances that these concerns come into existence. For that reason, a real ethical analysis is case-based and cannot be summed up in a list of principles or checklists.

Photo: Casu Canca, Founder and Director, AI Ethics Lab