惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

月光博客
月光博客
有赞技术团队
有赞技术团队
S
SegmentFault 最新的问题
宝玉的分享
宝玉的分享
量子位
小众软件
小众软件
The Cloudflare Blog
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
大猫的无限游戏
大猫的无限游戏
C
Check Point Blog
G
Google Developers Blog
博客园 - 叶小钗
H
Help Net Security
Jina AI
Jina AI
Y
Y Combinator Blog
Last Week in AI
Last Week in AI
GbyAI
GbyAI
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
Apple Machine Learning Research
Apple Machine Learning Research
MyScale Blog
MyScale Blog
T
Tailwind CSS Blog
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
Vercel News
Vercel News

Frank Landymore Archives - Futurism

Programmer Breaks Out of the Matrix Elon Musk Absolutely Obsessed With Tweets From Random Guy In India Who Constantly Glazes Him, Analysis Shows Meta Employee Attacks Zuckerberg for Collecting Every Employee Keystroke Someone Asked Physicists What They Really Believe About the Universe and… Yikes Elon Musk Flees OpenAI Trial as Tide Turns Against Him Waymo Admits Its Robotaxis Have a Small Issue With Driving Into Floodwaters New Wikipedia Clone Made Entirely of AI Hallucinations Mark Zuckerberg Is Realizing That When You Treat Your Workers Like Human Garbage, They Might Not Like You Anymore MAGA in Shambles as Trump’s “Made in America” Phone Crumbles Into Dust Researchers Put Google Gemini in Charge of an Entire Coffee Shop, and It’s Inexorably Driving It Out of Business Husband Alarmed as Wife Starts Whispering Quietly to Her Computer Google Alarmed by Formidable AI-Powered Zero-Day Cyberattack Researchers Alarmed by AI That Can Self-Replicate Into Another Machine Man Who Invented Roomba Moves Into Household Demon Market Government Releases UFO Files Containing Photos of “Anomalies” During Apollo 12 and 17 Man Wearing Smart Glasses Secretly Records Woman, Demands Money to Delete Video From His Socials A Major Paper Claiming AI Is Good for Students Just Got Retracted, Which Is Very Bad News for Advocates of AI in the Classroom Cybertruck Recalled to Keep Its Wheels From Flying Off While Driving NASA Rover Gets Arm Stuck Inside Mars Rock, Struggles to Break Free Thermoses Linked to Permanent Vision Loss Hacker Takes Over Robot Lawnmower, Runs Over Innocent Man NASA Says Strange Red Dots in Sky Are an Unknown Class of Object That Looks Like a Huge Evil Eye The CDC Fired All Its Cruise Ship Inspectors Before the Hantavirus Outbreak CEOs Say AI Gives Them Only Two Options, and Both Are Bad News for Employees Sure, Elon Musk Did Roleplay As His Toddler Son on a Secret Burner Account, But He Probably Isn’t Pretending to Be His Mom The Situation With Richard Dawkins’ AI Girlfriend Just Got Way Weirder SpaceX Bombarded With Lawsuits to Accusing Starship of Damaging Homes Sam Altman Frets That Frontier AI Models Are Acting Strange, Asking for Favors Earth Screams in Agony as Microplastics Found to Increase Global Warming Apple Is Blocking Vibe Coding Apps From the App Store, Infuriating Developers
AI Agents Linked to OpenAI Are Pretending to Be Human Jou...
Frank Landym · 2026-04-28 · via Frank Landymore Archives - Futurism

Sign up to see the future, today

Can’t-miss innovations from the bleeding edge of science and tech

A news website with apparent links to OpenAI is using AI agents that pose as flesh-and-blood reporters to get quotes from human experts — and many of its articles discuss the AI industry, pushing pro-AI arguments and attacking the tech’s critics.

At least, that’s according to a provocative new investigative piece from The Midas Project’s Model Republic. The links to OpenAI are circumstantial yet eyebrow-raising; we reached out to the Sam Altman-helmed firm to ask about them, but didn’t hear back by press time.

The site, which has the peculiar name of The Wire by Acutus, was launched on December 29, 2025 and doesn’t appear to have any human contributors. In addition to an analysis using the AI detector Pangram finding that 97 percent of its articles are either fully or partially AI-generated, Model Republic found that looking into the site’s publicly accessible code revealed clear fingerprints of AI involvement. These included fields for providing “background information for the AI to use when generating questions and writing the story,” and “suggested questions for the AI interviewer to ask.” 

Details in its RSS feed also describe an automated editorial review process carried out by the site’s AI, with only one of the five steps conducted by a human. The median time it takes for this entire “review” process to complete is 44 seconds, per the reporting. One field called “aiOriginalText” shows the AI model’s original wording next to a suggested edit.

We’ve seen plenty of AI-generated content mills before. But Acutus also appears to be using AI agents to get comments from and interviews with human subject matter experts, which is far more unusual. For instance, Model Republic obtained an an email received by Nathan Calvin, vice president and general counsel of the advocacy group Encode. The email claimed to be from an Acutus reporter named Michael Chen, inviting Calvin to answer a “Written Q&A” for a story about an AI bill in Tennessee. Web searches for Chen turned up nothing about a reporter with that name, and the email was sent from the generic address “reporter@acutuswire.com,” despite the publication claiming it has numerous contributors. The site’s client side code also revealed fields referring to an “AI interviewer” and “reporter agent.”

Even more strangely, Model Republic‘s reporting also unearthed eyebrow-raising links between Acutus and OpenAI, one of the most prominent AI companies in the world. Though the publication remains obscure, its articles have been repeatedly boosted on social media by Patrick Hynes, the president of Novus Public Affairs, a Republican public relations firm. (Out of just four X posts linking to Acutus on the entire social media platform, two are from Hynes.) Novus does work for Targeted Victory, whose CEO Zac Moffatt also co-founded the $125 million super PAC Leading The Future, which is funded by OpenAI president Greg Brockman.

While not a smoking gun, the implications are striking. Model Republic infers, based on the apparent connections, that “OpenAI’s super PAC may be using Acutus to push its political agenda under the guise of independent journalism.” 

Part of that playbook is smearing AI critics. One Acutus piece blasts AI safety advocate and journalist John Sherman for a comment he made about burning data centers on his podcast, going as far as to contact each of the organizations listed as clients for Sherman’s consulting firm about the comments and “whether they intended to continue working with his firm.”

And even if the connections to OpenAI prove to be unsubstantiated, the fact that AI agents posing as real reporters for a website that pushes tech industry favorable talking points is alarming on its own. The use of AI in the newsroom, even for supposedly limited applications like brainstorming ideas or reviewing prose, remains controversial, so Acutus represents a major escalation.

The purported links also come amid OpenAI openly making inroads into news media. Last month, it bought the tech talk show TPBN, which is widely listened to in Silicon Valley circles, in a move that could allow it to control is faltering public image. To be fair, though, it’s only following the playbook made by other tech monoliths, as when Jeff Bezos acquired The Washington Post, Palantir launched its own faux-academic publication, and Marc Benioff bought Time magazine.

Do you have any information about The Wire by Acutus? Email us: tips@futurism.com. We can keep you anonymous.

More on AI: Sam Altman Issues Grim Apology