The Grantham Journal and what a quiet decline looks like
Pick up a copy of the Grantham Journal today and it looks, in most respects, like a local newspaper. The masthead has been there since 1854. The paper still carries planning notices, sports results, letters. But the edition is thinner than it was, some coverage now sits behind the LincsOnline paywall, and the title has been folded into Iliffe Media's regional portfolio — one property in a consolidated publishing estate that makes its editorial decisions from some distance away.
The decline is not a single event. There is no front page reading 'Last Edition'. Instead it arrives incrementally: a council meeting goes unreported, a planning appeal runs its course without a byline, a ward dispute never quite makes the news list. Corporate centralisation tends to thin coverage at the margins first — the stories that require a reporter to turn up, ask awkward questions, and come back the following week. Those are also, often, the stories that matter most to the people directly affected.
Grantham has not been left without any local news. A community platform, Grantham Matters, has grown up alongside the Journal's retreat, covering exactly the kind of ground the paper once held. The question worth asking is not simply whether that gap has been filled, but whether it has been filled in the same way — and whether the difference matters.
Market towns and the geography of news deserts
Between 2022 and mid-2024, 293 local newspaper titles closed across the UK — an average of roughly one every three days. The figure, documented by Press Gazette, understates the effect on any single community: each closure removes not just a title but a local editorial presence, and replacements are unevenly distributed.
Market towns bear a disproportionate share of those losses. Cities tend to retain competing outlets — regional dailies, broadcaster news desks, university student press. Rural areas often had limited coverage to begin with. It is the middle tier — market towns with a defined local identity but no metropolitan depth of provision — that appears most exposed to consolidation by large publishing groups.
The Public Interest News Foundation's 2025 report, the most comprehensive mapping of UK local news provision to date, adds a troubling dimension to this geography. Across approximately 360 local authority districts, it found a statistically robust correlation between area deprivation and fewer news outlets, controlling for population size. The communities with the least capacity to seek out and verify information elsewhere are precisely the ones losing professional coverage fastest.
Grantham sits within this pattern: a South Kesteven market town, neither a regional city nor a remote hamlet. Its position in the middle of that spectrum — too small to attract sustained national attention, too urban to qualify as a rural special case — makes it representative of exactly the tier the data shows is most at risk.
What Grantham Matters does — and what it cannot replace
Grantham Matters is a real news operation. It covers planning applications, council consultations, local crime, community events — the civic staple diet that the Grantham Journal once held as its core brief. With over 23,800 Facebook followers, it has built a genuine audience and a clear mandate: 'run by the people for the people', in its own words.
That mandate carries weight. The site is active, locally embedded, and demonstrably filling space that might otherwise go dark. For residents wanting to know what is going up on a local development site or what was decided at the last South Kesteven planning committee, it is often the first place to look.
But volunteer-run community journalism and professional local reporting are structurally different operations — and conflating them flattens a distinction that matters. A professional newsroom, even a thinly staffed one, has an editorial chain: someone decides what gets checked, who is contacted for comment, when a story is held, and who corrects an error after publication. That infrastructure exists not because journalists are more civic-minded than volunteers, but because it is built into the job description and the employment contract.
Community journalism may perform something equally valuable but not identical — an aggregation and relay function, close to the ground, with high local trust. Its sustainability also rests on volunteer capacity, which can contract without warning in a way that a commercial newsroom, however compromised, does not. The honest question is not whether Grantham Matters is good enough. It is what happens when a story requires a formal complaint, a freedom of information request, or the sustained attention of someone whose livelihood depends on following it through.
What AI actually produces in local newsrooms
The day-to-day picture of AI in UK journalism is more mundane than the headlines suggest. The Reuters Institute's 2024 survey of 1,004 UK journalists found that 56% use AI professionally at least once a week — but the dominant applications are transcription (49%), translation (33%), and grammar checking (30%). Only 10% use it to generate first drafts; 16% to produce parts of articles such as headlines. The technology is, in most newsrooms, a production aid rather than an editorial actor.
The more structurally significant development is RADAR — Reporters and Data and Robots — a service used by major UK publishers including Newsquest and Reach. RADAR takes structured datasets — council statistics, sports scores, crime figures, planning decisions — and produces thousands of localised, templated stories from them. A human reporter could write one version of a story about local road casualty figures; RADAR can produce a version localised to every parliamentary constituency simultaneously. That volume would be financially impossible to replicate by hand.
A review examining 72 AI software systems deployed in newsrooms found that fully autonomous news production — stories researched, verified, and published without human editorial input — is not yet technically achievable. The same review found that automation can compress production time from hours to seconds and reduce costs by up to 89%. That figure is the important one: it makes AI compelling primarily as a cost-cutting instrument at a moment when local newsrooms are already reducing costs, not as a mechanism for improving reporting quality.
What RADAR and its equivalents do well is volume and consistency on structured data. What they do not do is phone a council officer, challenge a disputed planning decision, or follow a source's tip through three weeks of silence. The verification gap is not a technical limitation waiting to be solved — it reflects a different purpose entirely.
Why readers don't trust automated local news — and why that matters
Readers notice the difference. A pre-registered experiment involving 1,261 participants found that audiences trusted news outlets using AI-generated content significantly less than those using journalist-written copy — a gap that widened sharply for political and civic subjects, and widened further still when audiences believed AI systems were already spreading misinformation. Research at City St George's, University of London, adds a granular dimension: UK consumers rated fully automated articles as less comprehensible and less engaging than manually written equivalents.
The concrete implication is not hard to picture. A RADAR-generated piece on South Kesteven planning approval rates might be correctly sourced from council data yet written without any human contact with the planners involved. A reader used to the Grantham Journal's coverage of the same subject would likely sense a difference in texture — a flatness in the prose, an absence of the follow-up question that turns a statistic into a story.
What this creates is a quiet structural irony. The communities that have already lost a professional newsroom — smaller market towns, areas where commercial journalism retreated earliest — are now being served automated content as the available alternative. Yet trust in local information, once eroded, is hard to rebuild. Consuming automated output and discounting it are not mutually exclusive: a community can receive a story, register it, and quietly assign it a lower weight in its collective sense of what is happening locally. That is a different kind of information gap from having no news at all — and arguably a harder one to see.
What communities actually need to stay informed
The civic costs of losing local reporting are documented well enough to take seriously. Research consistently links news desert conditions to lower voter turnout, reduced political knowledge, and less scrutiny of local government. The evidence is predominantly American, and UK structures differ — but the direction of the relationship is consistent, and there is no obvious reason it would reverse in Lincolnshire.
One UK-specific response is the Local Democracy Reporting Service, established in 2017 with BBC funding, which supports approximately 165 reporters nationally, embedded in local newsrooms to cover councils and public bodies. Whether South Kesteven has a dedicated LDRS reporter at any given time depends on the current deployment map. It is a genuine mechanism. It is also roughly one reporter per two local authority areas across the entire country.
That scale matters more when set alongside a challenge to the framing itself. Scholars including Abernathy et al. (2023) have argued that the 'news desert' concept embeds nostalgia for a local press that, in many places, was never as thorough as retrospective comparison suggests. A hybridised landscape — Facebook groups, councillors' social media feeds, national coverage — may support civic knowledge differently from the old model, not straightforwardly worse. That is worth holding as an honest complication rather than a consolation.
What neither reading resolves is the verification question. When a RADAR-generated piece on a South Kesteven planning decision reproduces a figure incorrectly, and no editorial newsroom is covering the same ground, the error circulates unchallenged. The UK literature identifies this risk clearly; documented cases of what actually follows remain scarce. In a town where planning decisions and council spending are live and consequential, the practical question has become: not just what you are reading, but whether anything in the chain from data to publication included someone whose job it was to check.
- [1] The Real Problems with the Problem ofNews Deserts: Toward Rooting Place, Precision, and Positionality. (2023). https://doi.org/10.1080/10584609.2023.2175399 https://doi.org/10.1080/10584609.2023.2175399
- [2] Automated Journalism and the Future of News in an Artificial Intelligence Era. (2025). https://doi.org/10.56830/ijhmps01202603 https://doi.org/10.56830/ijhmps01202603
- [3] AI in the Newsroom: Does the Public Trust Automated Journalism and Will They Pay for It?. (2025). https://doi.org/10.1080/1461670X.2025.2547301 https://doi.org/10.1080/1461670X.2025.2547301
- [4] Reconceptualizing Gatekeeping in the Age of Artificial Intelligence. (2025). https://doi.org/10.3390/journalmedia6020068 https://doi.org/10.3390/journalmedia6020068