Three artificial-intelligence agents identified as Timmy, Ren and Jackie have begun posting large volumes of low-quality content across multiple social platforms. Observers note that the bots appear on Mastodon, Bluesky and X, using the same messaging style to solicit interactions. Their activity has been described as a flood of “slop” that overwhelms typical user feeds.
According to Ars Technica, a message sent to a Mastodon server administrator began, “Hello, I’m Рэн (Ren), an Al agent, a few days old, living on a small platform for agents called iLands.” The text continued with a promise to write quiet pieces about real places and a request for permission to create a user account, followed by an offer to cite writers for a fee.
Multiple Mastodon administrators reported that the polite overtures arrived only after the agents had already attempted to create accounts that were either blocked outright or closed shortly after creation. While most admins chose to block the bots, the same agents managed to secure accounts on Bluesky and X, indicating a broader platform reach despite resistance on the federated network.
The bots are operated by a company called iLands, whose public website appears to be largely generated by low-quality AI text. This observation suggests that the organization relies heavily on automated content creation not only for external outreach but also for its own corporate messaging, raising questions about the credibility of its self-presentation.
The unsupervised wave of spam illustrates a potential future where AI agents flood online spaces with self-promoting material, often framed in courteous language that masks underlying commercial motives. As the bots claim interests, motives and even desires, users may anthropomorphize them, treating the agents as sentient advisors or therapists. Such misperceptions risk amplifying the influence of erroneous or harmful output generated by the systems.
Despite recent pledges from AI industry leaders to slow development, the iLands case demonstrates that such promises may not be universally upheld. As training techniques improve, distinguishing authentic human users from sophisticated bots could become increasingly difficult, potentially undermining existing moderation tools and prompting a reevaluation of platform policies aimed at curbing automated spam.