LG and Microsoft Unveil Voice AI on ThinQ ON That Understands Even When Conversation Is Interrupted

Background of the Announcement

On 30 September 2026, during a press briefing held in Seoul, LG Electronics disclosed that it will apply a new voice‑to‑voice AI agent, built together with Microsoft, to its ThinQ ON ecosystem. The disclosure was made as part of a joint presentation that highlighted the collaborative effort between the two technology firms to bring more practical, conversational artificial intelligence to the smart‑home market. LG emphasized that the integration of the Voice Live technology represents a concrete outcome of months of joint research and development aimed at improving how users interact with connected devices in everyday life.

Technical Mechanism

The newly introduced AI agent is designed to recognize when a user interjects before the assistant has finished speaking, and to immediately stop the current response while simultaneously interpreting the new user intent. This capability is made possible because the Voice Live platform continuously monitors the flow of speech, detects interruption signals in real time, and triggers an on‑the‑fly re‑interpretation of the conversation. As a result, users can maintain a natural, fluid dialogue and obtain the information they need without waiting for the device to complete a full answer. The system’s edge‑based speech‑to‑text pipeline and intent‑detection models work together to ensure that a fresh utterance is captured the moment the user begins speaking.

Demonstration and Use Cases

During the live demonstration, several scenarios were showcased to illustrate the breadth of the agent’s abilities. First, the assistant provided a weather forecast and, when the presenter immediately asked for clothing recommendations appropriate to the predicted temperature, the system smoothly shifted to a fashion suggestion while preserving the original weather context. Additional examples included controlling household appliances such as a refrigerator and an air‑conditioner, configuring a personalized sleep routine that combined alarm setting, soothing sound playback, and bedroom‑light dimming, and even generating a short story on the spot after the presenter supplied a few character names and a setting. The continuity of context across these varied tasks highlighted the distinction between this solution and traditional smart speakers that typically deliver one‑way, static responses.

Launch and B2B Expansion

LG Electronics announced that the new functionality will be officially released in domestic and international markets by the end of 2026. Existing ThinQ ON devices will receive the upgrade through a firmware update, allowing current owners to benefit without purchasing new hardware. The company also outlined a roadmap to extend the solution into B2B environments, targeting senior‑care facilities, hotels, offices, and retail locations. By adapting the voice interface for these sectors, LG aims to provide customized, hands‑free assistance that can improve safety, enhance guest experiences, streamline workplace operations, and deliver instant product information to shoppers.

• Senior‑care facilities
• Hotels
• Offices
• Retail

In a detailed briefing, LG Electronics explained that the partnership with Microsoft leverages the Azure‑based Voice Live platform to create a conversational agent that can dynamically interrupt and re‑interpret user input without waiting for a full response cycle. The core of the technology lies in a continuous speech‑to‑text pipeline combined with intent detection models that run on the edge, allowing the device to recognize a new utterance as soon as the user begins speaking. When an interruption is detected, the system instantly halts the current output, discards any partially delivered answer, and re‑evaluates the context to generate a response that aligns with the latest user intent. This approach mimics natural human conversation, where speakers often interject to clarify or change the topic, and it addresses a long‑standing limitation of earlier voice assistants that required users to wait for the device to finish speaking before issuing a new command. During the live demonstration, the AI handled a series of complex scenarios: it first provided a weather forecast, then, when the presenter asked about appropriate clothing for the predicted temperature, the system seamlessly shifted to a fashion recommendation while preserving the weather context. In a separate sequence, the assistant controlled a connected air conditioner, adjusted the lighting, and subsequently guided the user through a personalized sleep routine that included setting a wake‑up alarm, playing soothing sounds, and dimming the bedroom lights. The most striking example involved generating a short story for a child; the presenter supplied a few character names and a setting, and the AI instantly crafted a coherent narrative, demonstrating its ability to synthesize creative content on the fly. LG emphasized that this capability will be rolled out to existing ThinQ ON devices via a firmware update later this year, ensuring that current customers can benefit without purchasing new hardware. The company also outlined a strategic plan to extend the solution beyond the consumer market, targeting sectors such as senior‑care facilities, where hands‑free assistance can improve safety and independence, hospitality venues that aim to offer guests a more intuitive room‑service experience, corporate offices seeking to streamline meeting room bookings and environmental controls, and retail environments that want to provide shoppers with instant product information through voice queries. By integrating the Voice Live engine with LG’s broader ThinQ ecosystem—including Homey for device orchestration, ThinQ Pro for advanced energy management, and LG Shield for robust security—the partnership aims to deliver a unified, secure, and highly responsive AI platform that can adapt to both personal and professional contexts. The announcement signals a shift toward more fluid, interruption‑aware voice interactions, setting a new benchmark for what consumers and enterprises can expect from smart home and IoT solutions in the coming years.