Three real scenarios from workplaces — study the scenario with your group, identify the violated or applied principles, then issue your verdict and recommendation.
Each group works on one scenario for 5 minutes — read the story, identify the affected principles, choose your verdict, and write one recommendation. Then your group presents its findings and discusses with the other groups.
A major consulting firm announced a "Business Analyst" position and received 2,000 applications in a week. The HR manager decided to use an AI tool to automatically filter the applications and shortlist only 20 candidates for interviews.
The system recommended its list in 3 minutes. No one from HR reviewed the remaining applications. The applicants did not know this system existed, nor did they have any way to know the reason for their rejection.
Later, one of the rejected applicants — an outstanding graduate — discovered that the system was giving lower scores to applicants who graduated from universities outside certain countries, even though they were internationally accredited.
A manager at a tech company struggles to evaluate the productivity of his remote team. He subscribed to an AI tool that connects to Microsoft Teams and automatically analyzes the number of messages, attendance times, response speed, and meeting participation — to produce a weekly "productivity report" for each employee.
The manager uses these reports in promotion and bonus decisions. The employees do not know this tool exists or that their conversations are analyzed. One of the top-performing employees received a low rating because they prefer phone calls over text communication.
A specialized hospital began testing an AI system to analyze X-ray images and predict diagnoses. The system was developed on one million X-ray images and proved 94% accuracy in laboratory studies — compared to an 87% average accuracy for human doctors in the same studies.
However, after 6 months of use, it was found that the doctors in the ER — due to workload — were approving the system's diagnosis in 80% of cases without reviewing the image themselves. In one case, the system misdiagnosed a cancerous mass, and the doctor approved it without review.