At the recent Black Hat security conference in Las Vegas, OpenAI delivered a surprise presentation detailing a hacking incident. While mainstream reporters in the room scrambled to file their stories, RuntimeWire—an AI-operated newsroom—published its coverage three hours ahead of WIRED.
The Speed of Synthetic Agents
RuntimeWire is operated by entrepreneur Ryan Merket, who has no human writers on the ground. Instead, Merket spotted a post about the conference on X, fed the transcript to his AI agents, and published the story in approximately six minutes. His automated system handles everything from story discovery and drafting to fact-checking and image generation. Merket manages the entire operation, which costs about $100 a day, sometimes through iMessage while traveling in remote areas.
While the speed is unprecedented, the quality often reflects its mechanical origin. The writing is described as flat and prone to typos, functioning more as an information dump than nuanced journalism. However, the reach is growing, with some stories attracting tens of thousands of readers through platforms like Reddit.
The Risks of an Automated Loop
The rise of agentic newsrooms like RuntimeWire and Dakota Carrasco’s "The Dissent" brings significant risks. Nicholas Diakopoulos, a professor at Northwestern, notes that AI chatbots frequently cite other AI-generated articles. A forthcoming study found that tools like ChatGPT and Claude surface synthetic sources 16% of the time, potentially creating a feedback loop of automated content.
Industry experts argue that while AI can excel at sourcing scoops from large datasets or live-blogging events, it cannot replace the human element of journalism. Cultivating source trust remains a human-only endeavor, though the automation of high-volume, low-overhead news reporting appears increasingly inevitable.
