Related Experiment Video
Updated: May 23, 2025

Characterization of the Sense of Agency over the Actions of Neural-machine Interface-operated Prostheses
Published on: January 7, 2019
Agentive linguistic framing affects responsibility assignments toward AIs and their creators
Dawson Petersen1, Amit Almor1,2
1Linguistics Program, University of South Carolina, Columbia, SC, United States.
Agentive language describing artificial intelligence (AI) influences responsibility perceptions. Lower AI experience increases perceived AI responsibility, while domain expertise mitigates this effect by reducing anthropomorphism.
Area of Science:
- Cognitive Psychology
- Human-Computer Interaction
- Linguistics
Background:
- Technology companies frequently employ agentive language to describe artificial intelligence (AI) capabilities.
- Linguistic research indicates that framing non-human entities as agents can lead readers to perceive them as causal actors.
- The impact of such framing on responsibility attributions towards AI and their creators remains under-explored.
Purpose of the Study:
- To investigate how agentive linguistic framing affects responsibility assignments to AI and companies.
- To determine if psychological anthropomorphism or linguistic causal schemas mediate these effects.
- To examine the moderating role of user experience with AI.
Main Methods:
- Participants read vignettes where an AI ("Dr. AI") provided harmful health advice, framed as either an agent or an instrument.
- Responsibility ratings were collected for the AI, the company, and patients.
- Participant AI experience was assessed.
Main Results:
- Participants assigned greater responsibility to the AI when it was framed as an agent compared to an instrument.
- An interaction was observed: lower AI experience correlated with higher AI responsibility in the agent framing.
- Higher AI experience participants showed no significant difference in responsibility assignment across framing conditions.
Conclusions:
- Agentive language increases perceived AI responsibility, particularly for less experienced users.
- Domain experience with AI mitigates the influence of linguistic framing by reducing anthropomorphism.
- Findings highlight the psychological impact of language on AI perception and accountability.
Related Concept Videos
Framing Effects
Language and Cognition
Fundamental Attribution Error
Attribution Theory
Stereotype Content Model
The Sense of Self: Reflected Self-Appraisal and Social Comparison

