Summarize Written Text

The regional broadcaster Merrow Public Media closed its last dedicated newsroom fact-checking desk in 2021, shifting verification duties onto individual reporters and a shared software tool that flagged statistically improbable claims for review. The decision followed a decade of shrinking advertising revenue and two rounds of layoffs that had already reduced newsroom staff by nearly forty percent. Management argued that the automated tool, trained on years of past corrections and wire service data, could catch most factual errors faster than a human deskmate previously could, freeing reporters to publish breaking stories with shorter delay. An internal review conducted eighteen months later produced mixed results. The average time between a story's submission and its publication fell by nearly a third, which newsroom leadership counted as a clear win in a competitive local media market where speed often determines which outlet's version of a story gets shared first. However, the same review found that the rate of published corrections nearly doubled compared to the final year the fact-checking desk had existed, with most new errors involving misattributed quotes, outdated statistics, and misidentified locations in photographs rather than the kinds of numerical anomalies the software was designed to catch. Younger reporters, who had less experience without a second set of eyes, accounted for a disproportionate share of the corrected stories. In response, Merrow's editorial board reinstated a scaled-down human review step specifically for stories involving direct quotations and visual material, while keeping the automated tool for numerical claims. Media researchers who studied the case have since cited it as an example of a broader pattern: automation tends to succeed at narrow, well-defined tasks it was explicitly trained for, but performs poorly on the more varied, judgment-based errors that experienced human editors previously caught almost incidentally.