With live news, every minute counts. Newsrooms covering a press conference or a breaking-news moment often hit the same bottleneck: someone has to watch the full stream, track the timings and transcribe the quotes that matter before a story, clip or alert can go out.
Reuters Live AI Transcription is built to remove that bottleneck. Now available on Reuters Connect, it provides automated, time-coded transcripts for selected English-language Reuters Live events, generated in real time as each event unfolds.
Find key moments faster
Time-coded transcripts let your team scan a live event and go straight to the quotes, soundbites and moments that matter, without scrubbing through the full video. The search for the right clip can start the moment the words are spoken.
Move from live to published at speed
By cutting the manual work of monitoring and transcribing, Live AI Transcription helps your team turn live coverage into articles, clips, alerts and social posts more efficiently, so you stay competitive while a story is still developing.
Built into the workflows you already use
Transcripts appear on the Reuters Live event page in Reuters Connect while the event is in progress and stay with the recording afterwards for replay and clipping. They are also available as an optional subtitle track (WebVTT) in the HLS streams delivered through Reuters Connect and the Reuters API, so teams that integrate Reuters Live directly into their own systems can use them too.
For those teams, delivery is getting more robust too: the Reuters API now offers SRT (Secure Reliable Transport), a new protocol for connecting to a Reuters Live stream with more reliable video delivery. Support on Reuters Connect will follow.
How it works
Transcription starts as a live, word-by-word feed that appears as the event is spoken, then updates to a refined, sentence-by-sentence version shortly afterwards. Transcripts are generated automatically using AI and are not reviewed by Reuters editorial; they are published live, as events happen, and are clearly labelled wherever they appear. Transcription is switched on for selected events, beginning with English-language news conferences.
This is just the start
Live AI Transcription launches with English and will expand to more of the languages spoken frequently across Reuters Live events, along with support for a wider range of events over time. It is one part of a wider investment in Reuters Video to help our customers work faster and more flexibly, from the live moment through to the finished story.
"A moment happening live and being able to highlight it for audiences is one of the real pressure points for our customers in today's rapid news cycle...Live AI Transcription is about closing that gap: a faster route from the moment something is said to the story, clip or alert our customers publish. This is the next step in our AI video journey, and we'll keep building on it."
Conor Molumby, Senior Video Product Manager, Reuters
