By GNA Team, GNA
Ho (V/R), Aug. 6, GNA – A simulation study has found that people are more likely to coordinate effectively with Artificial Intelligence (AI) when it is presented as a colleague rather than simply as a tool, highlighting the importance of trust in human-AI collaboration.
The study, presented by Dr Marwa Saab of the KAAF University at the International Conference on Artificial Intelligence in Healthcare and Pharma (ICAIH 2026) on Wednesday in Ho, examined how the way AI is framed influences coordination outcomes and whether those effects are driven by trust or role clarity.
Presenting a paper titled “AI as a Tool: A Simulation Study of Relational and Structural Mechanisms in Human-AI Coordination,” Dr Saab said the research was grounded in established theories, including the Computers as Social Actors Theory, Social Exchange Theory, Task-Technology Fit Theory and Role Theory.
She explained that people often interact with AI as though it were a social being because of its human-like characteristics, such as voice, personality and conversational abilities, while decisions to adopt particular technologies are also influenced by the perceived benefits they offer and their suitability for specific tasks.
The study involved 100 participants, with 56 assigned to an AI “colleague” condition and 44 to an AI “tool” condition.
The researchers assessed coordination outcomes using trust and role clarity as mediating factors, while AI familiarity was examined as a moderating variable.
The findings showed that participants who viewed AI as a colleague reported significantly higher levels of trust and achieved better coordination outcomes than those who viewed AI merely as a tool.
Conversely, participants in the tool category recorded higher levels of role clarity, suggesting that while defining AI strictly as a tool may improve understanding of responsibilities, it does not necessarily enhance collaborative performance.
Dr Saab said statistical analysis indicated that trust played a stronger role than role clarity in improving coordination between humans and AI, while AI familiarity also appeared to influence collaboration, although further research was needed to better understand its impact.
She stressed that the findings were based on a simulation and would need to be validated through real-world experimental studies before firm conclusions could be drawn.
According to her, the study suggests that successful human-AI interaction may depend more on how users perceive their relationship with AI than on clearly defined structural roles.
Dr Saab called for further research to examine how social framing influences clinicians’ perceptions of AI and how those perceptions affect trust, teamwork and decision-making in healthcare settings.
She said the findings could help guide the design and implementation of AI systems that support stronger collaboration between healthcare professionals and intelligent technologies while improving patient care and clinical outcomes.
GNA
Edited by Maxwell Awumah/Benjamin Mensah