Microsoft AI Agent:
Building trust in AI-powered meeting workflows

Our team is exploring how working professionals build trust with AI agents in workplace video conferencing. As AI becomes more autonomous, such as suggesting action items and drafting follow-up emails, users face a paradox: they either overtrust or distrust these systems. Through expert interviews and user research, we're designing interaction patterns that help people know when to trust, when to verify, and when to step back, enabling productive human-AI collaboration in everyday work.