OpenAI postpones launch of voice ChatGPT, AI still not safe

OpenAI announced on Tuesday that it was postponing the launch of the voice assistant features of its ChatGPT chatbot, citing the need for further security testing. The company had initially planned to introduce the new feature to some paying subscribers at the end of June, but decided to push back the initial release by a month. The new features will be available to all paying users in the fall, although the company stressed that the exact timing will depend on whether security and reliability standards are met.

Last month, OpenAI showed off what’s new in ChatGPT in a demo that sparked excitement among ChatGPT users. OpenAI introduced ChatGPT’s ability to speak in synthetic voices late last year, and the May demo used one of these voices to showcase a more advanced AI system called GPT-4o, capable of speaking in expressive tones, responding to a person’s tone of voice and facial expressions and have more complex conversations.

Amid the race among the world’s largest technology companies to compete in generative artificial intelligence, several projects have encountered unexpected obstacles. Last month, Google reduced how often it shows AI-generated answers at the top of search results, after the tool made bizarre mistakes such as suggesting putting glue on pizza. Last year, Microsoft made changes to its AI chatbot after it sometimes provided bizarre and aggressive responses.

OpenAI said Tuesday that it needs more time to improve the new voice version of its chatbot, making it better at noticing and blocking certain content, without revealing specific details. Many AI tools have been criticized for making up false information, spreading racist or sexist content, or showing bias in their output. Designing a chatbot that attempts to interpret and imitate emotions increases the complexity of its interactions, opening up new ways for potential errors. “ChatGPT’s advanced voice mode can understand and respond with emotions and nonverbal cues, bringing us closer to natural conversations in real time,” OpenAI said in its release. “Our mission is to bring these new experiences to the public with the utmost care.”