OpenAI delays ChatGPT ‘Voice Mode’ citing safety reasons

Last month, OpenAI unveiled GPT-4o and demoed a new advanced ‘Voice Mode’ for ChatGPT that can “understand and respond with emotions and non-verbal cues,” saying it will be available to a small group of people late June.

However, in a post on X, the company stated that they will be delaying the feature by a month because they are “improving the model’s ability to detect and refuse certain content.” It added that the new functionality won’t be available until it meets certain internal safety and reliability checks.