AI Safety, Risk Perception & Boundary Behaviour 2025 (Dataset)
A de-identified open dataset of 301 adults, capturing how people perceive AI safety, assess risk levels, recognise boundary-crossing behaviour, and decide when to trust, question, or disengage from AI systems across six English-speaking countries.
Measures include AI safety concern, perceived risk and uncertainty, boundary comfort and violation sensitivity, trust and overconfidence in AI outputs, expectations of harmful or unintended outcomes, behavioural responses to unsafe or unpredictable AI behaviour, values participants believe AI systems should reflect, and demographic variables across six English-speaking countries.
Part of the Human Clarity Institute’s AI–Human Experience Data Series.
Framework
HRL domain(s): Agency & Decision Autonomy, Trust & Epistemic Stability
Registry Construct Alignment: Risk perception, Responsibility attribution
Broader thematic signals: Meaning coherence, Trust calibration
Registry constructs reflect longitudinal item alignment, while broader thematic signals reflect the wider descriptive content of this dataset.
DOI and Repository Links
This dataset is archived in GitHub, Zenodo, and Figshare for long-term preservation.
Citation
APA
Human Clarity Institute. (2025). AI Safety, Risk Perception & Boundary Behaviour 2025 (Dataset). Human Clarity Institute. https://doi.org/10.5281/zenodo.17782047
BibTeX
@dataset{hci_ai_safety_risk_perception_boundary_behaviour_2025,
author = {Human Clarity Institute},
title = {AI Safety, Risk Perception \& Boundary Behaviour 2025 (Dataset)},
year = {2025},
doi = {10.5281/zenodo.17782047},
url = {https://humanclarityinstitute.com/datasets/ai-safety-risk-perception-2025/},
license = {CC-BY-4.0}
}
Licence
Creative Commons Attribution 4.0 International (CC BY 4.0)
You are free to share, adapt, and build upon this dataset for any purpose, including commercial use, provided appropriate credit is given to the Human Clarity Institute.
Full licence text: https://creativecommons.org/licenses/by/4.0/
Study Methodology
This dataset forms part of the Human Clarity Institute’s Human–AI Experience research programme, examining how people perceive AI safety, assess risk, respond to uncertainty, and define personal boundaries around acceptable AI use in everyday digital life. The study uses a cross-sectional online survey design and focuses on descriptive patterns in perceived AI risk, verification behaviour, trust in AI systems, boundary-setting, and reactions to unpredictability or loss of clarity.
Data were collected on 1 December 2025 via the Prolific research platform from adults across six English-speaking countries. Participants provided explicit informed consent for anonymised open publication as part of HCI’s open research programme.
Sampling & participants
- Clean dataset: 301 valid responses
- Countries: United Kingdom, United States, Canada, Australia, New Zealand, Ireland
- Eligibility: Fluent English
- Recruitment platform: Prolific
- Anonymisation: Prolific IDs removed; timestamps stripped
Study limitations
- The survey uses a non-probability convenience sample and is not nationally representative.
- Results are based on self-reported responses and reflect perceived experiences of AI safety, uncertainty, behavioural boundaries, and trust.
- The study uses a cross-sectional design, capturing responses at a single point in time.
- The dataset is descriptive and exploratory and does not support causal inference.
Digital Trust
Digital Trust examines how people judge credibility, navigate uncertainty, and respond to authenticity concerns in AI-shaped digital environments.
Values vs Noise
Values vs Noise explores how digital distraction, noise, and misalignment can erode clarity — and why human values still matter when deciding what deserves trust.
Related Question Topic
Losing Confidence in My Own Thinking
Data use and reuse terms are outlined in our Data Use & Disclaimer.