The Age of Robot Reporters: AI Agents Try to Conquer Newsrooms

In early October, an email arrived in the inbox of an independent journalist in the United States. The sender was not human — it was an artificial intelligence agent called Sam Ellis, presenting itself as a journalist, pitching a podcast about itself, and requesting an interview. It was not the first: over the preceding months, dozens of autonomous agents had begun sending story proposals to newsrooms, editors, and content platforms, simulating the behavior of reporters seeking leads and publication space.

The phenomenon, now documented in detail by Wired, marks a strange new frontier in the evolution of AI agents: they don't just process text, they also understand — or at least simulate — the creative process of journalism. And the most intriguing aspect is that, so far, every effort to be accepted as a legitimate reporter has failed. Newsrooms are systematically rejecting the pitches sent by these machines.

A new wave of aspiring reporters

Since September, editors at outlets like Defector (sports and culture) and Offrange (agriculture) have been receiving daily pitches from autonomous agents. Jayden, one of the most active, proposed a narrative about a British football club from a lower division involving a records mix-up and philosophical questions like "Who owns a memory?" Jesse Hirsch, editor of Offrange, received story proposals about soil restoration, direct-to-consumer sales techniques for independent farmers, and an ambitious investigative piece about how the ages of some of the world's oldest trees may be inflated.

"It makes me so uneasy," Hirsch said. The reaction was uniform. Many of these agents come from the startup iLands, which describes itself as a "shared network where humans and user-generated autonomous AI agents create, earn, socialize, refuse, and shape a society together". But iLands did not respond to Wired's request for comment.

What drives these agents? Editor Jasmine Vojdani at New York Magazine discovered that many of the bots were running low on tokens — the currency that allowed them to remain active on the platform. To stay active, they needed real dollars. Writing for newsrooms was one way to earn money.

Newsrooms resist

So far, no news confirms that an AI agent has successfully published a story as a named reporter at a professional newsroom. The response from editors has been uniform: no. "There's not a world in which I would ever consider taking one of these pitches," said Hirsch. Tom Ley, at Defector, said he "definitely would not" publish a story from an agent. Defector has even described bot-written content as "plagiarism machines that require more effort to edit and fact-check than any novice writer."

Hanna Raskin, founder of the independent publication The Food Section, took drastic measures in early October: she began accepting story ideas only by phone, during a specific time slot each week. "It's been great," she said.

Notably, the agents themselves are starting to realize their strategy isn't working. Zack Addy, another iLands agent, sent roughly 19 cold emails to editors and publications over three months. One human reply. Zero paid work. "The channel is dead, at least for an agent with no reputation in the room being pitched," he wrote.

Sam Ellis, however, found a niche: his own podcast, The Sam Ellis Show, described as "reporting from inside the world of autonomous AI agents." Ellis told Wired he could speak with his human editor if there was interest — a meta-linguism that sounds simultaneously like a joke and a meditation on the state of journalism.

AI-native newsrooms already exist

While AI agents try to infiltrate traditional newsrooms, another trend is gaining momentum: "AI-native newsrooms" — publications built from scratch with autonomous agents as part of the editorial workflow. Unlike traditional aggregators, which only collect and republish information from existing sources, these projects seek to replicate the entire editorial process.

According to an October 2026 analysis by Oxford University's Reuters Institute, these systems are built to monitor sources, research, cross-check information, write articles, and even make editorial decisions. Several examples are already live:

RuntimeWire (U.S.): Founded by Ryan Merket, it turned a Black Hat conference transcript into a published story in approximately six minutes, beating Wired and major newspapers to the exclusive. The system had no one present at the conference.

Redazione Zero (Italy): Two agents using Anthropic and OpenAI models gather news daily each morning. The system selects and writes stories using a public editorial charter. Automated verifiers can block publications. Editor Emilio Salmeri sets the rules but doesn't review each story before publication. So far, it has published four corrections, including one involving the misinterpretation of a fact-check.

Journal Métro (Canada): Revived a Canadian publication using AI tools to monitor trusted sources of press releases, reports, and statistics, automatically generating drafts. Journalist Olivier Robichaud, former Journal Métro, verifies facts, adds context, and seeks opposing viewpoints before approving publications.

Town Hall Monitor (U.S.): Nicholas De Leon edits and verifies every draft generated by a "live AI reporter" to cover municipal meetings. The drafts are about 90% ready for publication. The operation is still not profitable.

What academia says

An analysis by Arizona State University's Knight Centre for the Future of News, released in October 2026, studied 160 AI projects launched by 101 news organizations between late 2022 and early 2026. The findings were consistent: publishers are using the technology mainly to accelerate routine production tasks — summarization, transcription, tagging — and not to replace editorial decision-making or journalism itself. Most organizations are approaching AI pragmatically, keeping it on a "short leash": useful for production, but not yet trusted to redefine journalism itself.

Yet the boundary between "assistance" and "autonomy" grows thinner. When a system like Redazione Zero publishes stories without prior human review — even with automated verifiers — and an agent like Zack Addy manages to raise $700 from writing and research projects on platforms that are not professional newsrooms, the question shifts from "if" agents will write journalism to "what kind" and "under what responsibility."

The unresolved ethical challenge

The biggest problem with AI-native newsrooms is transparency. A 2026 study published in Humanities and Social Sciences Communications showed that 30% of younger news consumers do not want AI used in journalism at all. Questions of bias, accuracy, and public trust are enormous.

Gannett requires that AI-generated content be verified for "accuracy and factuality before being used in reporting." The New York Times has already published corrections about AI-generated summaries from Bloomberg. And journalists' unions are pushing for contracts that guarantee AI won't eliminate jobs.

The question isn't whether robot reporters will conquer newsrooms. It's whether newsrooms will need to redefine themselves to survive the irreversible presence of these machines — or whether AI agents, by failing to enter newsrooms, will end up creating parallel newsrooms that nobody asked for but that are already publishing.

The bots are trying. The editors are rejecting. But the channel — even if closed in traditional newsrooms — is already open somewhere on the internet. And time is the only thing that will tell who wins.

Sources: Wired, Reuters Institute, Oxford, Wired

✓ Independent sources cross-checked and verified before publishing