OpenAI’s latest report has highlighted a unique concern: users of ChatGPT’s new Voice Mode might be forming emotional bonds with the AI chatbot. The findings, part of the “GPT-4o System Card” safety review, suggest that prolonged interaction with the AI’s human-like voice and emotive capabilities could lead users to seek companionship from the chatbot, potentially reducing their need for human interaction.
The report also raises concerns about the AI’s influence on social norms, noting that extended use of Voice Mode could subtly alter users’ behaviour. While the report addresses various safety challenges, including the risks of generating inappropriate content, the potential for users to develop emotional attachments to the AI is a particularly noteworthy risk. OpenAI plans to continue studying this phenomenon, recognising that the emotional reliance on AI might require new safety measures. Currently, the only recommended solution is to limit the time spent using the Voice Mode.
