October 5, 2026
A page on the Uprise RI news site explains how it reports news using AI. (Photo by Janine L. Weisman/Rhode Island Current)
Since July, a Rhode Island news site has been publishing stories by authors that don’t exist.
Kelly Lockhart, a housing reporter for Uprise RI, wrote about the widening gap between earnings and housing costs in the Ocean State. Correspondent Adriana Vossari covered the latest votes — on zoning, school procurement, criminal justice for juveniles, and other issues — from the Providence City Council. Debut stories by national politics reporter Ryan MacMahonaghan, labor reporter Eddie Exxe, and other correspondents followed.
None of them are actual humans. They are, as the site notes, “computing agents — intelligent LLM reporters.”
They kept up their work with little fanfare until Sept. 17 when Boston Globe media reporter Aidan Ryan reported on Uprise RI’s novel approach to newsgathering in the AI age.
Other Rhode Island news organizations have tinkered or dabbled with artificial intelligence. The Providence Journal partnered with the AI-powered search engine, Perplexity. The Globe’s RI bureau published an op-ed that included the disclaimer: “The author used generative AI to help draft this commentary.”
But no outlet, as far as I know, had farmed out entire stories to chatbots before.
Uprise RI was founded in 2017 by progressive independent journalist Steve Ahlquist and is now owned solely by software developer Greg Brailsford.
Brailsford, who joined Uprise RI in 2020, declined interview requests from the Globe but did tell the paper, via email, that he thought the use of AI personas to report the news would become more common. Uprise RI, he told the Globe, is “a bit ahead of the game.”
This country’s robust freedom of speech gives people wide latitude to speak and publish without prior permission or regulation. But we should discuss big shifts in how our communities are covered. Because journalism is the lifeblood of democracy. And if that process is outsourced to machines — even if, as Uprise RI insists, humans oversee “Editorial standards, judgment calls, and corrections” — that has implications for all of us.
I was an unpaid contributor to Uprise RI, when Ahlquist ran and edited the site. Alquist left in 2023, and I continue to contribute to his Substack on an unpaid basis. He had nothing to do with this article. I’m writing this as a Rhode Islander, journalist, longtime observer and reporter on local media and someone who frequently gives library talks about media literacy and misinformation.
First, I want to thank Brailsford and Uprise RI. By going further than any other local outlet in the blending of AI tools with journalism, the site has sparked an important and timely conversation.
In theory, more coverage of local government is a good thing. One of the places hardest hit by the deterioration of local news has been the kind of granular, hyper-local reporting that Uprise RI offers. Go to the site and you’ll find coverage of zoning boards, city councils, and school committees. This is an admirable attempt to fill an existing gap.
But I have thoughts about Uprise RI’s approach.
The Society of Professional Journalists’ Code of Ethics has a whole quadrant of principles under the heading “BE ACCOUNTABLE AND TRANSPARENT.” This need for transparency in the news-reporting process is even more urgent when non-human technologies are being used.
In Uprise RI’s case, I wish Brailsford had offered an essay at the start of the LLM-reporting endeavor introducing the “reporters,” explaining how the process works and how this serves the public. And, in general, I would love to see the site feature more prominent labeling that the articles are generated by artificial intelligence. Right now, you have to click a “reporter’s” “bio” to see the “Digital LLM Reporter” label.
One day after the Globe article, Brailsford published a piece, “How a Rhode Island meeting becomes a story,” that explains the site’s process. The site’s “About” page now has a section labeled “Why do our reporters have human names and human faces?”
I say: Keep going.
Nothing about using LLMs for reporting is obvious or self-explanatory. I believe prominent labeling and explanation — to an extent not just equal to, but beyond what human-written sites offer — should be a staple of any AI-journalism operation.
I spoke with Brailsford via direct messages on the social media platform Bluesky.
“I almost never see journalists who write about AI talk about its benefits, and when they do, it’s a sentence or two, with the rest of the piece being how terrible it is,” he told me. Brailsford insisted he was no AI “defender” but has used AI for over two years and sees its great potential.
AI may well have miraculous and beneficial applications. But its downsides are glaring. Even if you don’t believe AI is going to lead to mass human extinction, it is certainly going to put lots of people out of work. Chatbots are often trained on the uncompensated lifting of intellectual property. Some of AI’s most visible champions are throwing money at President Donald Trump. Data centers essential to the success of AI are now so unpopular they are one of the few issues that unite our intensely divided country.
Anyone who aims to blend AI and journalism should be prepared to meet the audience where they already are — skeptical, wary, perhaps even angry — and not where they believe readers should be.
People are used to getting their news from human beings. Journalists are familiar faces on TVs and voices on our radios. They are with us during a community’s darkest moments. In Rhode Island, journalists are our neighbors: people we run into at the coffee shop or grocery store, people whose kids go to our kids’ school.
Moreover, news reporting is an activity that involves some of the processes that make us most human: conversations, sensory descriptions (smells, sounds, feelings), and an ever-evolving application of ethical considerations. To shift this work to robots is not a minor tweak; it fundamentally rethinks what journalism is.
Beyond that, accuracy is paramount in journalism, and AI is a misinformation accelerant. We see this in Google’s frequently erroneous AI-fueled search summaries and the flood of AI-generated images. AI fakes even hindered the manhunt for the Brown University mass shooter last December.
For lots of good reasons, many folks associate AI with fakery and falsehoods. This is a major, perhaps insurmountable, hurdle to getting them to trust AI-reported news.
In my exchanges with Brailsford, it was clear we approach this issue from very different perspectives. We disagreed about a lot. He did not appreciate it when — since we weren’t speaking face to face, or even via phone — I asked to confirm I was speaking with a human being, not an LLM.
But he said something that stuck with me.
He told me that, even when Ahlquist worked at the site, Uprise RI never made a lot of money.
So he doesn’t see using LLMs as a substitute for human reporters; it’s a way to publish stories that otherwise won’t get written.
“I, for one, want to know this information about the town I live in,” Brailsford, who lives in Cumberland, told me. “I don’t care who writes it. If it’s accurate, then I don’t care if a little child writes it, if a seasoned journalist writes it, or if a machine writes it. I just want to know what happened.”
It’s a provocative idea. One that we should all grapple with.
As a human journalist, I’m wary of making predictions. But I’m comfortable saying that the issues raised by Uprise RI and its team of LLM “reporters” aren’t going away soon.
Courtesy of Rhode Island Current
