The Risks of Using ChatGPT to Obtain Common Safety-Related Information and Advice

Oviedo-Trespalacios, O., Peden, A. E., Cole-Hunter, T., Costantini, A., Haghani, M., Rod, J. E., … Reniers, G. (2023). The risks of using ChatGPT to obtain common safety-related information and advice. Safety Science, 167, 106244. https://doi.org/10.1016/j.ssci.2023.106244

Posted: 16 Feb 2023 Last revised: 31 Jul 2023

See all articles by Oscar Oviedo-Trespalacios

Oscar Oviedo-Trespalacios

Delft University of Technology; Delft University of Technology - Faculty of Technology, Policy and Management

Amy E. Peden

University of New South Wales (UNSW) - School of Population Health

Thomas Cole-Hunter

University of Copenhagen - Section of Environmental Health

Arianna Costantini

University of Trento - Department of Psychology and Cognitive Science

Milad Haghani

University of New South Wales (UNSW)

J.E. Rod.

Universidad del Norte

Sage Kelly

Queensland University of Technology

Helma Torkamaan

Delft University of Technology

Amina Tariq

Queensland University of Technology

James David Albert Newton

University of Queensland - School of Psychology

Timothy Gallagher

Utrecht University

Steffen Steinert

Delft University of Technology - Faculty of Technology, Policy and Management

Ashleigh Filtness

Loughborough University

Genserik Reniers

Delft University of Technology

Date Written: February 3, 2023

Abstract

ChatGPT is a highly advanced AI language model that has gained widespread popularity. It is trained to understand and generate human language and is used in various applications, including automated customer service, chatbots, and content generation. While it has the potential to offer many benefits, there are also concerns about its potential for misuse, particularly in regards to providing inappropriate or harmful safety-related information. To explore ChatGPT's capabilities in providing safety-related advice, a multidisciplinary consortium of experts was formed to analyze nine cases across different safety domains: using mobile phones while driving, supervising children around water, crowd management guidelines, precautions to prevent falls in older people, air pollution when exercising, intervening when a colleague is distressed, managing job demands to prevent burnout, protecting personal data in fitness apps, and fatigue when operating heavy machinery. The experts concluded that there is potential for significant risks when using ChatGPT as a source of information and advice for safety-related issues. ChatGPT made incorrect or potentially harmful statements and emphasized individual responsibility, potentially leading to ecological fallacy. The study highlights the need for caution when using ChatGPT for safety-related information and expert verification, as well as the need for ethical considerations and safeguards to ensure users understand the limitations and receive appropriate advice, especially in low- and middle-income countries. The results of this investigation serve as a reminder that while AI technology continues to advance, caution must be exercised to ensure that its applications do not pose a threat to public safety.

Keywords: Artificial Intelligence, Responsible risk management, Risk communication, Safety Science, Chatbot, Human-AI Interaction

Suggested Citation

Oviedo-Trespalacios, Oscar and Peden, Amy E. and Cole-Hunter, Thomas and Costantini, Arianna and Haghani, Milad and Rod, J.E. and Kelly, Sage and Torkamaan, Helma and Tariq, Amina and Newton, James David Albert and Gallagher, Timothy and Steinert, Steffen and Filtness, Ashleigh and Reniers, Genserik, The Risks of Using ChatGPT to Obtain Common Safety-Related Information and Advice (February 3, 2023). Oviedo-Trespalacios, O., Peden, A. E., Cole-Hunter, T., Costantini, A., Haghani, M., Rod, J. E., … Reniers, G. (2023). The risks of using ChatGPT to obtain common safety-related information and advice. Safety Science, 167, 106244. https://doi.org/10.1016/j.ssci.2023.106244 , Available at SSRN: https://ssrn.com/abstract=4346827 or http://dx.doi.org/10.2139/ssrn.4346827

Oscar Oviedo-Trespalacios (Contact Author)

Delft University of Technology ( email )

Stevinweg 1
Stevinweg 1
Delft, 2628 CN
Netherlands

Delft University of Technology - Faculty of Technology, Policy and Management ( email )

P.O. Box 5015
2600 GB Delft
Netherlands

Amy E. Peden

University of New South Wales (UNSW) - School of Population Health ( email )

Thomas Cole-Hunter

University of Copenhagen - Section of Environmental Health

Nørregade 10
Copenhagen, DK-1165
Denmark

Arianna Costantini

University of Trento - Department of Psychology and Cognitive Science ( email )

Rovereto, Trento 38068
Italy

Milad Haghani

University of New South Wales (UNSW) ( email )

Disability Innovation Institute
UNSW, Randwick
Sydney, 2052
Australia

J.E. Rod

Universidad del Norte ( email )

Barranquilla
Colombia

Sage Kelly

Queensland University of Technology

Helma Torkamaan

Delft University of Technology

Amina Tariq

Queensland University of Technology

James David Albert Newton

University of Queensland - School of Psychology ( email )

Australia

Timothy Gallagher

Utrecht University

Steffen Steinert

Delft University of Technology - Faculty of Technology, Policy and Management ( email )

P.O. Box 5015
2600 GB Delft
Netherlands

Ashleigh Filtness

Loughborough University ( email )

Ashby Road
Loughborough, LE11 3TU
Great Britain

Genserik Reniers

Delft University of Technology ( email )

Do you have negative results from your research you’d like to share?

Paper statistics

Abstract Views
4,053
PlumX Metrics