TIME has a history of embracing new ways to deliver its journalism to a global audience. Continuing this tradition, TIME is launching an AI-generated audio briefing to make its reporting more accessible. This tool, developed with ScaleAI, aims to reach audiences where they are.
The new audio briefing utilizes generative AI to transform TIME’s news articles into a discussion between two AI bots named Henry and Lucy, a nod to TIME’s founder, Henry Luce. The goal is to make TIME’s journalism more digestible and interactive by delivering trusted and accurate reporting in a conversational manner. The audio briefing will only discuss content previously published by TIME.
This tool is another avenue for distributing TIME’s journalism, which is already available on TIME.com, social media, in print, and through TIME Studios’ film and television projects. This audio briefing marks further progress in TIME’s use of AI for news delivery. Following the launch of TIME AI, the company plans to continue updating and improving its AI products, with the aim of enhancing accessibility and engagement with TIME’s journalism.
Here’s a deeper look at its functionality.
How does the audio briefing work?
The tool selects news stories from TIME’s newsletter, The Brief, which is curated by a TIME editor. It employs GPT/o models, a family of large language models from OpenAI, a TIME strategic partner, that can generate content in formats beyond text. The GPT/o models from OpenAI were used for script generation, and GPT4o-mini-TTS was used to convert the script into an audio file. This process is fully automated and uses only stories published on TIME.com.
What voices are used in the audio?
Henry and Lucy have unique voices developed in partnership with ScaleAI, using OpenAI’s voice library. These voices were tuned to align with TIME’s editorial style and accurately reflect the original reporting. The AI hosts are designed to properly credit both TIME reporters and the sources cited in The Brief.
What safeguards are in place?
The AI models are programmed to base their discussions on the original articles and always attribute information to its source. They are not allowed to deviate from the article content. As noted in the audio player, there is a possibility of error.
How else does TIME use AI?
In December 2024, TIME introduced an AI toolbar to answer questions about TIME’s Person of the Year, drawing from TIME articles, other reliable sources, and the tool’s existing knowledge. This tool was designed to provide accurate information while maintaining a professional conversation. More information about the TIME AI toolbar is available here.
“`