The Glasgow Times published a story this week about a car crash in Possilpark that ended with an editor's note left visible in the text: "If you want, I can go back and rework the earlier RTC/fire copies in this thread to match that exact quoting style." The article has since been updated, but the mistake reveals how deeply AI-generated copy has embedded itself into the city's local newsrooms.
Newsquest, the UK's second-largest regional news publisher, owns the Glasgow Times, the Herald, the National, and several smaller local titles. It now employs 36 AI-assisted reporters across its publications through an initiative that produces roughly 9,000 stories a month. The company argues the technology frees journalists from routine tasks to focus on "high-impact reporting that matters to readers and communities."
When Newsquest first launched the scheme, readers could identify AI-assisted stories by clicking through to a reporter's bio. That disclosure has since disappeared. At Reach, the UK's largest regional publisher, which owns the Daily Record and Glasgow Live, a system called Guten scrapes stories from other Reach publications and rewrites them to match house style. No disclaimer accompanies the output.
The speed economics of automated news
Lewis James, senior data scientist at Reach, and Dan Taffler, director of data and analytics, wrote a blog post explaining that Guten handles "ingestion of news wires, wire article suggestions, article idea recommendations, and content generation." They reported the system cut breaking news publishing time from nine minutes to 90 seconds.
Both publishers have cut journalism jobs over several years while scaling up AI output. The economics are straightforward: AI is cheap. Humans ask for decent wages, take holidays, and require time off. The market, as the source material puts it, "will give people what they demonstrate they want, and if that's dirt-cheap journalism, that's exactly what they'll get."
What the machines cannot do
AI cannot have sensitive conversations with devastated parents outside a school entering liquidation. It cannot stand at the scene of a car crash and notice what the police report omits. It cannot channel community grief into reporting that captures what went wrong and why it matters. These tasks require a reporter on the ground, hours of research, and sometimes weeks of work.
For writers and journalists watching this shift, the implications are direct. The technology that publishers frame as a productivity tool is also a cost-cutting mechanism. The same systems that rewrite wire copy and generate neighborhood news summaries are being deployed at outlets that have been reducing headcount. This is not theoretical - it is already happening across more than 200 regional titles in the UK alone.
Why this matters for writers
If audiences won't pay for human-generated reporting, publishers will supply whatever is cheapest to produce. The result is more automated content, more aggregation between sister publications, and fewer working journalists. The counterweight is reader-funded models that employ reporters, pay illustrators and photographers, and publish work that requires being physically present. For writers, the career paths are diverging: toward volume-driven AI-assisted production on one side, and toward subscriber-supported original reporting on the other. The speed at which this split widens depends on whether readers fund the second option.
Your membership also unlocks:
