OpenAI has unified engineering, product, and research teams over the past two months to overhaul its audio models in preparation for an audio-first personal device set to launch in about a year, according to reporting from The Information.
The overhaul targets improvements in OpenAI’s audio capabilities beyond current limitations. The company’s new audio model, scheduled for release in early 2026, will produce more natural-sounding speech. It will manage interruptions in a manner similar to a real conversation partner. This model will also enable the AI to speak simultaneously while the user talks, a function that existing models cannot perform.
OpenAI plans a family of devices powered by this advanced audio technology. Possible designs include glasses or screenless smart speakers. These devices aim to function as companions rather than mere tools, integrating seamlessly into daily interactions.
Former Apple design chief Jony Ive contributes to OpenAI’s hardware initiatives. OpenAI acquired his firm io for $6.5 billion in May. Ive prioritizes audio-first design principles to address device addiction. As The Information notes, Ive views this approach as an opportunity to “right the wrongs” of past consumer gadgets.
The push toward audio interfaces aligns with broader industry developments. Smart speakers, featuring voice assistants, exist in more than a third of U.S. homes. These devices have established voice interaction as a standard household element.
Meta introduced a feature for its Ray-Ban smart glasses that employs a five-microphone array. This setup assists users in hearing conversations within noisy environments. The technology effectively transforms the glasses into a directional listening device positioned on the face.
Google initiated experiments in June with “Audio Overviews.” This feature converts search results into conversational summaries delivered via audio. Users receive spoken overviews instead of visual lists, facilitating hands-free information access.
Tesla incorporates Grok and other large language models into its vehicles. The integration creates conversational voice assistants capable of natural dialogue. These assistants manage tasks such as navigation and climate control through spoken commands and responses.
Startups pursue similar audio-centric hardware with mixed outcomes. The Humane AI Pin, a screenless wearable, consumed hundreds of millions in funding before emerging as a cautionary tale in the sector. The Friend AI pendant functions as a necklace that records aspects of users’ lives while providing companionship. This device has generated privacy concerns alongside reports of existential dread among users.
Additional startups develop AI rings projected for debut in 2026. Sandbar represents one effort. Another involves Pebble founder Eric Migicovsky. These rings enable wearers to engage in conversations directly with the device on their hand. Form factors vary across these projects—wearables, pendants, rings—but all emphasize audio as the primary interface. Spaces such as homes, cars, and even the face evolve into interactive audio environments.
