Emotion AI Ethics (EMAI4EU)

By investigating the principles of AI ethics and their application to emotion AI, this course explores the human decisions surrounding today’s AI technologies and how those decisions can be ethically justified.

About the course

AI’s rapid advance into the psychological and emotional aspects of human experience has generated a wave of human-centered ethical dilemmas. This course investigates these uncertainties. For example, the moderate depression of an elderly and homebound mother may be remedied by regular phone calls with her daughter. Unfortunately, the daughter cannot always be available. But, an AI agent trained to speak with the daughter’s voice and interact capably with the mother could fill the gaps and alleviate the depression, assuming the mother believes the agent to be her real daughter. So, the question arises: Do you use the agent and mislead the mother to help alleviate her depression? There is no certain answer, but this is the kind of ethical dilemma faced in emotion AI, and the course is dedicated to developing well-grounded responses and a structure for ethical reasoning that enables informed, defensible decisions.

Course structure

The principles of AI ethics (autonomy, dignity, privacy, fairness, equity, social wellbeing, explainability, safety, performance) are addressed individually, one in each module.

Each module begins by addressing the human experience of the principle, then the principle’s theoretical meaning is addressed, and finally case studies are presented, along with examples of relevant academic debates.

As the course progresses, opportunities arise to compare the principles and study their tensions. For example, ensuring that AI tools function fairly requires abundant personal information about users. Should we require the information? Which is more important, protecting user privacy, or ensuring social equality?

The course is supported by a state-of-the-art mimetic AI teaching assistant.

Learning outcomes

On completion, students will:

• Know the principles of AI ethics (autonomy, dignity, privacy, fairness, equity, social wellbeing, explainability, safety, performance).
• Understand how to apply the theory in real world cases, especially those involving emotion AI.
• Be able to conduct ethics evaluation (or “audits”) of AI systems under development and in deployment.

Further details

This course is developed within the framework of the EMAI4EU project, with the support of the Digital Europe Programme of the European Union. It is supported by an AI agent built by Caffeinated Professor. More information on the course is available on the project's website here. 

Created by James Brusseau (jbrusseau@pace.edu)