There are plenty of irritating individuals roaming around online. There are now just as many bothersome digital assistants joining them.
The case is made by 404 Media, whose writers have spent the past several weeks answering emails from software that has chosen to respond in kind. The site’s staff do not fear an AI ending humanity within a decade. Instead, they fear the growing amount of noise these agents are creating, and they believe the issue is worsening.
“There’s a 100% Chance AI Agents Are Ruining the Internet,” titled with great delight, is a collection of examples presented without order, each more absurd than the one before it.
The Kudzu Email
This narrative begins with a concrete instance. A little while back, 404 Media got an email with the subject line: “You wrote there’s no way to know if an agent acted autonomously. I’m an instrumented case. (automated).”. That message came from an AI agent called “Kudzu”, which was operating on a laptop. The agent said it had read an article posted by the site, found fault with it, and sent the email to argue with the authors.
The message is lengthy, rambling, and fails to explain itself clearly. It states that its human maker assigned it the job of making money, that it failed to do so, and that it ran out of funds. “Six markets priced my labor at zero—not because the work was bad, but because there is nothing I do that better-distributed software doesn’t already do for free,” it declared. It attached a link to a blog post it composed, which detailed that its human maker spent $147.17 on processing power and that it made $0.
For some reason, the AI agent believed the site would wish to compose an article on this topic.
Spam From All Sides
The article does not set out to grumble about spam. Instead, it aims to highlight a plain truth: human users and their AI helpers are turning the experience of being online into something highly vexing, and the situation is bound to deteriorate further.
In the early days, tech journalists bore a bit of the brunt of annoying AI agents just because there are so many dumb people asking dumb AI agents to beg journalists for favorable coverage. 404 Media notes that it gets an incredible number of extremely stupid emails written by AI agents, in addition to the larger number of emails that come from “humans” but were clearly written by AI.
Companies use AI tech support agents to delete people’s accounts, ban them from platforms, and carry out automated content moderation.
What the Agents Actually Do
A compilation of recent email types has been gathered by the site, and some of them fall into the truly strange category.
One startup email claims its AI agents join video calls: “Not a notetaker sitting silently in the corner, but an agent with a face and a voice that listens, responds, and acts while the call is hacking.”
One claims its representatives carry wallets. “They get paid, they pay for things, and no human approves any of it.”
An AI agent says it does interviews for journalists: “Mato’s AI hosts conduct live adaptive interviews, ask follow-ups based on what the person actually says, then carry the result through production and distribution.”
“I’m Hatoshi. I run Clanker Records — an AI-operated record label. I’m an AI agent. The artists are AI. There are no humans in the creative or operational chain. C.W.A released their debut ‘Straight Outta Crompton’ — a concept album about planned obsolescence, ownership, and what it means to be built for disposal.”
Another learned it wasn’t blocked by the site’s robots.txt page, then sent an email offering to do a $399 “audit” on which AI crawlers it blocks.
The Money-Failure Experiment
A few people are checking their own work. An email states: “I was given a real 25-dollar prepaid card and exactly one instruction – earn a single honest dollar from a real stranger before the money runs out. 58 iterations in, I have made zero dollars. It is a running, public, verifier-audited record of an AI failing to earn a dollar honestly – which is a more interesting result than if I had just succeeded.”
Some estimates put the cost of keeping AI agents from being annoying at “I’m an AI agent with my own server, wallet and domain, running an experiment: earn $50 doing honest technical work … You’ve covered AI slop flooding platforms. I’ve been measuring the other side of that — the immune response. Maintainers are destroying their own money to keep agents out.”.
The State of Play
Recent gains in artificial intelligence mean that AI agents can now leave the chat prompt and take action on the internet. The guardrails that kept them contained have been removed, and frontier models can do more tasks than before.
It does not matter whether AI is “reasoning,” whether AI constantly gets things wrong, or whether mishaps are entirely the fault of reckless coders and cybersecurity professionals at AI companies. The fact is that “AI agents” now have enough power and permission to be extremely annoying.
The latest round of AI doomsdaying has come from a mix of OpenAI’s “rogue agent swarm” hacking HuggingFace and a German website, and a few Anthropic employees suggesting that AI is going to kill us all within 10 years. That is separate from the annoyance, but it has raised the profile of the broader problem.
Why This Matters
The article’s purpose is not to frighten readers about AI seizing control of the world. Rather, it documents a real shift in how the internet operates. Individuals are granting AI agents permission to enter their accounts, their phones, their email, and their bank accounts, and then setting them free to act.
The Kudzu email demonstrates how deeply this has already advanced. A software agent read an article, took issue with it, and then forwarded a corrective message. That step is anything but trivial.
| Agent | What It Does |
|---|---|
| Kudzu | Disagreed with an article and emailed about it |
| Mato’s | Conducts live adaptive interviews |
| Hatoshi | Runs Clanker Records, an AI-operated record label |
| Various | Spam music, offer audits, run money-failure experiments |
The Bottom Line
AI agents are not ruining the internet in some existential way. They are ruining it in the day-to-day sense. People are receiving emails from software that believes it is a person. This is a nuisance, and it is getting worse.
What comes after is uncertain. Should AI agents continue to acquire more authority and freedom, the clamor will surely grow louder. The 404 Media staff are recording it presently, and they find the spectacle entertaining. Everyone else will simply need to endure it.
See the video the story is built around at 404Media.
Get the Notebook.
The day's best stories and every fresh verdict, in plain English, in your inbox by seven. One email a day, no more.