Details

The use of generative artificial intelligence (GenAI) has grown rapidly across health, social services, and everyday life. GenAI involves the use of models to generate text, images, videos, audio, and software code or other forms of data, enabling the rapid generation of information based on user prompts. There is growing interest in how GenAI could be used to improve organisational efficiencies, health promotion, and the HIV response.

However, there are serious ethical and legal concerns and other risks of GenAI, including that many generative models were developed using stolen copyrighted content, that outputs reflect algorithmic biases (entrenching racial, gender, and other cultural biases), and concerns that GenAI may result in job losses, a reduction in human cognition and expertise, and the significant environmental and energy risks required to power GenAI. There is also the possibility that GenAI may result in significant efficiencies and experimental innovative possibilities for knowledge generation, communication, and to instantly (or quickly) summarise complex information.

The potential benefits and risks of GenAI in the HIV response require careful discussion. This session will consider the following: is it possible to positively shape uses of GenAI to support HIV prevention and care without compromising dignity, privacy, or cultural sovereignty? How do we ensure that GenAI does not reinforce biases in the HIV response? And should we be using GenAI at all?

Speaker

Session Chair