Cleveland Paper Uses AI to Write Articles

The Cleveland Plain Dealer, a 184-year-old newspaper, is using artificial intelligence to draft articles on topics ranging from ice carving festivals to medical research. These AI-assisted articles are identified by the byline "Advance Local Express Desk" and include a disclaimer at the end stating the piece was produced with AI tools and reviewed by staff. The move is part of a broader strategy by the paper's parent company, Advance Local, which operates in 10 U.S. markets. The company is partnering with the Swedish firm United Robots to automate data-driven content like real estate sales and weather alerts, a strategy that has reportedly led to significant traffic increases in markets like Michigan (174% rise) and Oregon (68% rise). Editor Chris Quinn has defended the practice, arguing it frees up reporters from writing to focus on gathering information, effectively adding an extra workday each week. He stated the goal is to restore "hyperlocal" coverage that was lost when the paper closed suburban bureaus over a decade ago, with reporters now expected to file four AI-assisted stories daily. The implementation has been met with backlash from some journalists and industry veterans. A prospective reporting fellow withdrew their application upon learning the job involved no writing, only filing notes for an AI tool. Critics have expressed concern that relying on AI will cause news to lose the essential human nuance that helps communities understand why events matter. This push into AI comes as the local news industry faces a severe crisis, with projections showing the U.S. will have lost a third of its newspapers and nearly two-thirds of its newspaper journalists between 2005 and the end of 2024. While some newsrooms are using AI for tasks like transcribing interviews and brainstorming headlines, most maintain a policy against AI generating entire articles. The Plain Dealer's approach involves a human "rewrite specialist" who directs the AI, with reporters getting the final review before publication. This human-in-the-loop system is a key safeguard, as studies have shown AI assistants can misrepresent news content, including factual inaccuracies, up to 45% of the time.