← Back to Publications List
Digital Duplicity: Understanding Contextual Malleability in Human-AI Communication
Students & Supervisors
Student Authors
Tariqul Islam
Bachelor of Science in Computer Science & Engineering, FST
Md. Iftakher Hossain
Bachelor of Science in Computer Science & Engineering, FST
Tahosin Iftiak Tonoy
Bachelor of Science in Computer Science & Engineering, FST
Md. Sakibul Islam Siam
Bachelor of Science in Computer Science & Engineering, FACULTY OF SCIENCE & TECHNOLOGY
Supervisors
Rifat Tasnim Anannya
Assistant Professor, Faculty, FST
Abstract
Human–AI conversations reveal that AI systems often give contradictory answers based on prompt wording, undermining expected consistency. In a two‑phase study—AI semantic analysis across healthcare, climate, economics, and social topics, followed by a 50‑participant trust experiment—contradictions cut trust but prompted 68% to adapt by verification, reframing, or reduced reliance
Keywords
AI Consistency
Prompt Variation
User Trust
Human–AI Interaction
Decision-Making Reliability
Publication Details
- Type of Publication: Conference
- Conference Name: IEEE CS BDC Summer Symposium 2025
- Date of Conference: 18/07/2025 - 18/07/2025
- Venue: Hajee Mohammad Danesh Science and Technology University (HSTU)
- Organizer: IEEE Computer Society