OpenAI's newest AI ambassador? Someone's grandma

share on

twitter /
facebook /
linkedin /
- email
- telegram
- whatsapp
- wechat
- pinterest
- line
- snapchat
- reddit

OpenAI has enlisted an unlikely group of ambassadors to introduce its latest voice AI technology: grandmothers.

In a nearly four-minute campaign film launching GPT-Live, three grandmothers put the new voice experience through its paces, showcasing how it can hold more natural conversations while helping with everyday tasks ranging from trip planning and fact-checking to real-time language translation.

The campaign was designed to demonstrate GPT-Live's conversational improvements in a way audiences could immediately experience rather than simply be told about, Brandon McGraw, interim chief marketing officer, consumer and hardware at OpenAI, told MARKETING-INTERACTIVE.

Don't miss: Did OpenAI just make Messi the GOAT of prompting?

"The new model can listen, pause, respond with better timing and handle the natural rhythm of conversation, including interruptions and back-and-forth, so we wanted a creative idea that would demonstrate those qualities immediately," McGraw said.

He added that grandmothers offered a "warm, fun and universally recognisable" way to bring ChatGPT Voice to life, noting that they are confident conversationalists who naturally tell stories, interrupt and ask follow-up questions, making them an ideal way to show whether GPT-Live could keep pace with real conversations.

The film opens with one grandmother knitting a sweater as she introduces GPT-Live, describing it as a new generation of voice models designed to make conversations with AI feel more like talking to another person. She then asks the AI to explain the technology in simpler terms before passing the conversation to another grandmother.

Throughout the film, the women interact with GPT-Live conversationally rather than issuing commands, highlighting its ability to respond naturally while assisting with tasks such as planning a holiday to Santorini, verifying information for a lecture and translating French in real time. The campaign closes with all three grandmothers reuniting to deliver one final endorsement of the new feature.

According to McGraw, the creative was also intended to broaden the appeal of ChatGPT Voice beyond existing users by making the experience feel approachable rather than technical.

"Instead of leading with model architecture, we show what the breakthrough means for an everyday person, which is speaking naturally and being understood," he said.

The campaign accompanies the launch of GPT-Live, OpenAI's latest voice model built to enable more fluid, human-like conversations with ChatGPT.

Unlike previous voice systems that waited for users to finish speaking before responding, GPT-Live uses a full-duplex architecture, allowing it to listen and speak simultaneously. The model can acknowledge users with cues such as "mhmm" or "yeah", pause while they gather their thoughts or continue listening without interrupting, creating a more natural conversational flow.

OpenAI said GPT-Live also separates real-time conversation from more complex reasoning tasks. When users ask questions requiring web search, deeper reasoning or agentic capabilities, the voice model delegates those requests to GPT-5.5 in the background while continuing the conversation, before returning with the completed response. The architecture also enables live translation and allows the voice experience to continuously benefit from OpenAI's latest frontier models as they are released.

While these capabilities can sound abstract in technical language, McGraw said the campaign was designed to make the improvements immediately tangible.

"The grandmother film is a demonstration of it. It shows the model navigating the spontaneity and unpredictability of a real conversation, which helps people feel the difference rather than asking them to take our word for it," he said.

The launch also introduces an upgraded ChatGPT Voice experience, with more natural conversations, improved listening capabilities, smarter responses and visual answer cards for topics such as weather, stocks and sports. Users can also choose between different reasoning levels depending on whether they prioritise speed or more considered responses.

OpenAI said GPT-Live has undergone expanded safety testing covering areas including self-harm, emotional reliance, violence and sexual content. The company has also introduced voice-specific safeguards, including protections for teen users, monitoring for emotionally sensitive interactions and measures designed to prevent voice impersonation.

GPT-Live is rolling out globally across iOS, Android and ChatGPT.com. GPT-Live-1 will become the default voice model for Go, Plus and Pro subscribers, while GPT-Live-1 mini will power the experience for free users.

McGraw added that while GPT-Live is rolling out globally, including across Asia Pacific, OpenAI has no local creative adaptations to announce at this stage. However, multilingual experiences and regionally relevant storytelling will remain an important part of the company's continued expansion of ChatGPT Voice across different markets.

The launch of GPT-Live comes amid a broader slate of product announcements from OpenAI. Alongside the new voice experience, the company unveiled its GPT-5.6 family of models, comprising GPT-5.6 Sol, Terra and Luna, designed to improve performance and efficiency across coding, reasoning, scientific research and knowledge work.

OpenAI also introduced ChatGPT Work, an AI agent that can independently complete multi-step tasks such as creating presentations, documents, spreadsheets and marketing briefs by working across connected apps and workflows.

Separately, Microsoft said GPT-5.6 will become the preferred model powering Microsoft 365 Copilot across Word, Excel, PowerPoint, Chat and Cowork. The update is aimed at helping users create, analyse and collaborate more effectively within Microsoft's productivity suite using OpenAI's latest flagship models.

share on