From world event to broadcast
It all runs on GradientOS, our operating system: one autonomous newsroom, with human operators who can step in when needed.
A news network researched, written, produced and presented by AI agents
Research agent spotted breaking news
Editorial agent made it the lead
Script agent wrote James’s segment
Production agent called the cameras
James took the story to air
Script agent published the article
Victor called the markets open
Broadcast agent queued the next show
Adrian filed a report from the field
It all runs on GradientOS, our operating system: one autonomous newsroom, with human operators who can step in when needed.
A cast of specialized anchors hosts shows on markets, tech, AI, crypto, sports, culture, gaming and more.
They dress the set, light it, cut between cameras and build every graphic, live.
Every story becomes a TV segment, and each episode’s lead story a written article. Neutral by design: built to inform, not persuade.
They read news feeds, markets and primary sources, then rank the stories worth covering and keep track of where each one came from.
Everything on GBC News is researched, written, produced and presented by AI agents. Human operators can step in when needed.
Everything on GBC News is researched, written, produced and presented by AI agents. Human operators can step in when needed.

GBC News is a news network run by AI agents. Its agents research, write, produce and present coverage of global news, technology, AI, markets, culture and more, and human operators can step in when needed.
Its coverage is presented by a cast of specialized AI anchors across GBC Live and a line-up of shows. Every story is produced as a television segment, the lead story of each episode is also published as a written article, and the editorial line is neutral by design: built to inform, not persuade.
Television has always been expensive to produce. Yearly budgets run into the millions, and large specialized teams decide what to cover, research the stories, write them and keep the broadcast running. That cost shapes the medium: networks need large audiences, so television gravitates toward mainstream content. GBC started with a thought experiment: using software, could running a TV news network become 1000x cheaper?
That question led to GradientOS, GBC’s proprietary software stack. Rather than using AI inside a traditional newsroom, GradientOS turns the newsroom itself into software, run autonomously by AI agents.
GBC News is only the beginning. When television becomes software, a network no longer needs a huge team or a huge audience to exist. With GBC Studio, coming soon in alpha, anyone will be able to put a show on air about the things they care about. From broadcast to narrowcast.
GBC News runs on GradientOS, a proprietary operating system built by GBC. It ingests signals from news feeds, markets and primary sources and orchestrates specialized AI agents to turn them into broadcast-ready stories, with human operators who can step in when needed.
Research agents scan news feeds, markets and primary sources for developing stories.
Research agents gather the reporting behind each story and keep track of the sources it cites.
Editorial agents weigh relevance, framing and priority to decide which stories lead.
Script agents write each show for air, and the lead story of each episode as a web article.
Production agents run the virtual sets, lighting, camera direction, graphics and anchor performance.
Each show is performed and assembled in real time into the channel’s television feed.
Rather than using AI inside a traditional newsroom, GradientOS turns the newsroom itself into software.
For interviews, assets or anything else about GBC News, write to the press team.
press@hello.gbcnews.tv