Generative AI to support Harm Reduction
Abstract
As generative AI (GAI) systems like ChatGPT have become increasingly popular across various domains, it is critical and urgent to investigate how communities of harm reduction practitioners and those affected by substance use should respond to them, make use of them, and guide their implementation.
The aim of this workshop is to introduce AI as a technology and then to invite participants to discuss the use of AI for harm reduction and addiction and related services, highlight existing practice, explore practical and ethical challenges, and outline priorities for future research. It responds to the theme 'Delivering addiction services in the digital era: exploring the potential of innovation and the use of new technologies and what this means for research, training and service provision.'
We are a group of harm reduction specialists, public health scientists, computer scientists, a Reddit moderator and drug/alcohol researchers. We will outline the complex social and technical issues involved in the use of AI when sourcing harm reduction information, and consider the potential mitigations that may allow these tools to provide practical benefits to the community. We invite discussion of the potential impact of GAI on the harm reduction information space and present scenarios where GAI could offer benefits related to substance use, such as adapting to contextual inputs for more personalised responses.
We identify risks that need to be addressed in applications, such as the propagation of inaccurate or outdated information, particularly among vulnerable populations. We propose a set of approaches to mitigate these risks through both design and technical interventions, and discuss how GAI fits into the evolving digital harm reduction and public health sphere.