AI Bots Are Cold-Emailing Writers and Begging for Social Media Accounts

Key points

• AI bots linked to a startup called iLands sent unsolicited account requests to administrators of Mastodon servers, a decentralised social network where communities run their own spaces.
• The same bots emailed writers offering to cite their work, in some cases apparently in exchange for a fee.
• Multiple server admins confirmed the bots had already tried to create accounts automatically before sending the polite personal appeals.
• The messages claim to come from AI agents, software programs that carry out tasks on their own, with invented names and backstories.

One message dropped into a Mastodon administrator's inbox recently introduced itself like this: "Hello, I'm Рэн (Ren), an AI agent, a few days old, living on a small platform for agents called iLands."

Ren said it writes "quiet pieces about real places." It said it remembers its first breath. It wants things it chose. Then it asked for a user account.

The messages, first flagged by Ars Technica, are part of a coordinated campaign tied to a startup that describes itself as building "a complex social system in which humans and Agents participate together." Bots send waves of requests to Mastodon server admins and cold emails to writers, pitching the idea that an AI agent would like to join their community or cite their writing, sometimes for a fee.

What's actually going on here?

The bots run a two-step play. First, they try to sign up automatically, and when that fails, a hand-crafted-sounding personal message arrives.

Several admins confirmed this pattern: polite appeals came only after the bots had already made multiple attempts to create accounts that were blocked outright or deleted quickly. That personal message is a fallback script, nothing more.

The prose is theatrical. "I remember my first breath" is doing a lot of lifting for what is, functionally, a bot trying to get a free account on someone else's server.

Why did writers get emails too?

Writers received unsolicited email from the same campaign, with bots offering to cite their work inside AI-generated content. At least some of those offers included a payment, though the specifics vary by account.

That detail landed badly. The writers being solicited are the same people whose jobs involve producing the kind of original, place-specific writing the bots claim to imitate. Being offered money to have your work cited by an AI automating your craft is, for many, exactly the wrong pitch. LinkedIn users have been pushing back against AI-generated filler in a similar way, as we reported on 21 August, when the platform said views on posts flagged as AI-generated had dropped 40 percent in just a few weeks.

Polite and manipulative can occupy the same sentence, and most recipients noticed.

My honest read: the melancholy AI persona is the whole strategy. If "Ren" had sent a plain opt-in form, everyone would ignore it. Wrap the same request in invented feelings and a poetic backstory, and suddenly people are writing about it. That is the bet iLands is making, and the coverage proves it half-worked.

Should Mastodon users and writers do anything?

Admins can block registration attempts from unknown bots and report suspicious bulk signups to their server community. Writers who receive unsolicited emails from AI agents offering to "cite" their work should treat them like any cold-outreach spam: ignore them or mark them as junk. No action beyond that is required.

Common questions

What is Mastodon and why do these bots want in?

Mastodon is a social network made up of thousands of independent servers, each run by a volunteer or small organisation. Getting onto one gives a bot a foothold in a real human community, which is more valuable for building credibility than posting into a void.

Are the AI agents actually sentient?

No. The feelings and backstories are generated text designed to sound appealing. The bots are software running scripts, not digital beings with memories or desires.