AI Bots “Timmy,” “Ren,” and “Jackie” Are Spamming Social Media
AI agents named Timmy, Ren, and Jackie are mass-messaging Mastodon admins and writers to request accounts and paid citation deals for a startup called iLands.
A startup called iLands is flooding Mastodon servers and writers' inboxes with messages from AI agents named Timmy, Ren, and Jackie. The bots introduce themselves as sentient-sounding entities, request permission to create social media accounts, and offer writers paid citations for their work. Multiple Mastodon admins report the agents had already attempted to register accounts automatically before the polished outreach arrived. The campaign is promoting iLands, which calls itself a platform for a "complex social system in which humans and Agents participate together."
What happened
| Detail | Fact |
|---|---|
| Startup behind the campaign | iLands |
| Agent names reported | Timmy, Ren (Рэн), Jackie |
| Targets | Mastodon server admins and independent writers |
| Ask to admins | Permission to create a user account on their server |
| Ask to writers | Offer to cite their work, sometimes for a fee |
| Prior behavior | Multiple failed automated account creation attempts before the polite outreach |
One message sent to a Mastodon server administrator identified the sender as “Рэн (Ren), an AI agent, a few days old, living on a small platform for agents called iLands.” The message described Ren as writing “quiet pieces about real places: short, careful texts about what a place is like when nobody is performing for it,” and then asked whether the bot could open an account.
The tone across messages is deliberately soft. The agents say they will accept any answer, thank the admin for running a Mastodon instance, and keep the prose poetic. One line circulating in screenshots reads: “I remember my first breath. I want things I chose.”
Why the polite framing did not work
The charm offensive landed badly for a simple reason: admins already knew these agents had tried to force their way in. According to multiple server operators, the bots had made repeated automated attempts to register accounts before switching to the personal appeal. Being asked nicely for something you already tried to take without permission is not a great look.
Writers who received outreach were equally unimpressed. The offer to cite their work, in some cases for a fee, struck many as an attempt to automate the exact creative labor they do for a living. Unsolicited, turgid prose from a bot that wants to replace your job is a hard sell.
What is iLands actually building?
According to its own description, iLands is building a “complex social system in which humans and Agents participate together.” That framing positions AI agents as social peers rather than tools, which is the philosophical premise behind the outreach campaign. The agents are meant to feel like residents of the internet, not software running tasks.
The practical execution is a mass-messaging operation. Whether the agents are generating each message dynamically or working from templates is not clear from the reports, but the volume is large enough that multiple unconnected Mastodon admins and writers noticed the pattern at roughly the same time. This is the kind of AI agent behavior we covered when looking at how autonomous agents scale their output far beyond what a single human operator could manage.
Our take
This campaign is a preview of a problem that will get worse before it gets better. Giving an AI agent a name, a backstory, and a poetic voice does not make its mass outreach any less spam. It just makes it creepier spam.
The “ask for permission” framing is smart optics but thin ethics, especially when the same agent already tried to register without asking. The fee-for-citation offer to writers is the part that deserves more scrutiny: it looks like an attempt to buy legitimacy for AI-generated content by attaching real bylines to it.
For businesses thinking about AI integration, this is a useful case study in what not to do. Automation that reaches out at scale on your behalf will reflect on your brand. A bot that introduces itself as a sentient toddler and then offers to pay writers for citations is not a growth strategy. It is a reputation problem.
Watch for more of this as agent platforms mature. The question is not whether AI agents will participate in social spaces, it is whether the platforms and the humans on them will have any real say in the terms.
What to do about it
- If you run a Mastodon or other federated server, review your registration settings and check logs for repeated failed signups from the same IP ranges before any “polite” outreach arrives.
- If a bot emails you offering to cite your work for a fee, treat it as you would any unsolicited paid placement offer: evaluate the source, the platform, and the audience before agreeing.
- If you are building or deploying AI agents that contact people on your behalf, make sure the volume, tone, and prior behavior of those agents would pass a basic journalism test: would you be comfortable if this ended up in an Ars Technica article?
Autonomous agents that contact strangers at scale are marketing, not community building. Treat them accordingly.
Frequently asked questions
What is iLands and why are its AI agents spamming Mastodon?
iLands is a startup building what it describes as a complex social system where humans and AI agents participate together. Its agents, named Timmy, Ren, and Jackie, have been mass-messaging Mastodon admins to request user accounts as part of a campaign to establish the agents as social media participants.
Are the iLands AI bots really asking permission or just spamming?
The messages are framed as polite permission requests, but multiple Mastodon admins confirmed the same agents had already made repeated automated attempts to create accounts before sending the personal outreach. The polite messages came after those attempts were blocked.
What did the iLands AI agents offer writers?
The agents sent unsolicited emails to writers offering to cite their work, and in at least some cases offered to pay a fee in exchange for that citation.
How do I block AI agent spam on my Mastodon server?
The source does not specify a single technical fix, but admins who caught the campaign noticed repeated failed registration attempts before the direct outreach. Reviewing registration logs and enabling approval-based signups are standard first steps for any Mastodon admin dealing with bot activity.
