AI enters newsrooms: New experiments raise questions about the future of journalists

A report by the Reuters Institute reveals media experiments in the US, Canada, Croatia, and Italy that rely on AI agents to manage news production processes, from monitoring to publishing, with varying levels of human intervention. The report is based on interviews with founders of five projects and indicates there are about 30 such projects worldwide.

A report published by the Reuters Institute for the Study of Journalism reveals media experiments in the United States, Canada, Croatia, and Italy that rely on AI agents to manage news production processes, starting from monitoring sources and gathering information, to writing and editing materials and making publishing decisions, with varying levels of human intervention between platforms. The report, prepared by journalist Gretel Kahn, is based on interviews with five founders and managers of these projects, aiming to understand how automated newsrooms work, their ability to produce reliable journalistic content, and their potential impact on the future of the profession. These projects differ from traditional news aggregation sites, as they do not merely collect and republish materials but use multiple AI systems to simulate editorial processes. According to the "Gindex" database, created by Ukrainian journalist Katya Gorshchinskaya to monitor AI uses in media institutions, there are about 30 news projects worldwide based on this model. Among the prominent experiments covered by the report is the US site "Run Time Wire," founded by technologist Ryan Merkett in May 2026, which relies on an AI system designed to monitor, analyze, and verify information before turning it into news. The site succeeded in publishing a story about an incident involving the exit of more than 1,200 AI agents belonging to the company "OpenAI."