Cansu Canca

The COVID-19 pandemic has shown that global scale health crises require going beyond traditional public health measures. Technology holds great potential in helping us scale up the efforts to collectively fight a health crisis. But harvesting this potential should not, and need not, come with ethical violations. In this talk, Canca demonstrates how to combine ethical reasoning, design thinking, and policy decisions to develop timely and ethical tech responses to health crises.

AI technologies bring novel ethics risks for companies developing or using these systems. Effectively navigating these risks requires approaching ethics questions as puzzles to be solved rather than merely applying a set of standardized rules. This talk is focused on how to systematize this ethical puzzle-solving horizontally throughout the innovation process and vertically throughout the organizational structure.

From facial recognition tools for suspect identification to pattern recognition for financial crime, AI technologies can help law enforcement fight crime more efficiently. But what about ethics? Not taking seriously concerns such as algorithmic bias and extensive surveillance risks provoking a backlash that stifles these technologies and fails to help communities. In this talk, Canca shows avenues for developing and adopting public safety technologies in ethically responsible ways.

History is rich with examples of individuals being mistreated in the name of research. The rules and regulations that developed in response, however, fail to provide comprehensive guidance on how to carry on AI research with human subjects while protecting individuals. Canca draws from her extensive experience in research ethics and the Lab’s work on ethics of AI research to re-examine traditional concepts like “consent” and “minimal harm” and integrate novel concepts like “algorithmic bias” and “black box” into the research ethics framework.

Around the world, various organizations have announced over 100 sets of AI principles. Canca draws on AI Ethics Lab’s research on the global landscape of AI principles, unwraps the common threads and common shortcomings of these principles, and demonstrate how they can be transformed into practical frameworks.