OpenAI has announced a delay in the rollout of its advanced Voice Mode for ChatGPT, initially planned for late June. The company now aims for a launch in the fall after refining the feature.
“We need one more month to reach our bar to launch,” OpenAI tweeted. The delay is attributed to improvements in content detection and refusal, user experience enhancements, and infrastructure preparations to support millions of real-time responses.
OpenAI cited safety and reliability as key factors in the postponement, echoing a Bloomberg report on the need to address safety issues. The updated timeline depends on meeting high safety standards, with a limited alpha test group slated to receive the feature first.
Voice Mode is part of the new GPT-4o model, offering a more humanlike interaction. It can sing, tell jokes, translate languages, and narrate stories. “ChatGPT’s advanced Voice Mode can understand and respond with emotions and non-verbal cues, moving us closer to real-time, natural conversations with AI,” OpenAI stated.
Additionally, OpenAI plans to share more about its upcoming video- and screen-sharing capabilities, which will allow users to show ChatGPT their work for real-time assistance.
The delay follows OpenAI’s decision to remove the “Sky” voice from its voice feature after concerns arose about using actress Scarlett Johansson’s voice without permission.
