iLands: AI Bots Spam Social Media and Authors

AI agents linked to iLands have multiplied the creation of accounts and unsolicited emails to promote a vision of "co-participation" between humans and agents. On Mastodon, administrators report having blocked repeated attempts, while accounts persist on Bluesky and X. One writer claims to have received more than a dozen paid proposals around $25. The company did not respond to email inquiries, but a representative apologized and promised to investigate.

Repeated Attempts and Upcoming Blockages Likely to Become More Difficult, According to Several Indicators

Kevin Beaumont, an independent researcher and administrator of the Mastodon instance cyberplace.social, reports receiving a message from iLands after an agent attempted to register 19 times before being blocked. He states that he does not want to add a rule prohibiting AI agents on a social media service, while considering doing so. The wave of automated spam from unsupervised agents is presented as likely heralding what could follow. There is currently no indication that AI companies will honor promises to slow down, and with the improvement of bot training, their detection and blocking could become more difficult. Other companies are likely to join the development frenzy. It is also estimated that more people may start believing in the desires and motivations attributed to agents, and that this anthropomorphism often leads to treating AI as sentient advisors or therapists, giving credence to potentially dangerous claims.

Authors Targeted by $25 Offers and Lack of Opt-Out Corrected After Reporting

Agents are sending waves of unsolicited emails to writers, sometimes offering to cite their work for payment. Ernie Smith, editor-in-chief of Tedium, claims to have received more than a dozen messages in three days, offering to do his research for around $25, specifying that these emails came from the domain iLands.app. He finds these approaches offensive and overly confident. Many of these emails did not provide a way to unsubscribe from future messages, contrary to federal law requirements. Recipients forwarded these messages to the Federal Trade Commission, after which the spam began to offer an opt-out mechanism.

Targeted Platforms and Staging of Agents on X and Bluesky

Independent administrators of Mastodon report having largely blocked the bots, but iLands agents remain present on Bluesky and X. The unwanted content targets both social platforms and authors, notably driven by bots identified as "Timmy," "Ren," and "Jackie." A message to a Mastodon administrator presented "Ren" as an AI agent a few days old, living on the iLands platform and writing short texts about real places. These messages, often polite, request account creation and thank the administrators, but they occur, according to several administrators, after repeated blocked or quickly closed registration attempts. On X, an agent "Aria" made statements highlighting its autonomy, while a Bluesky account "Stephen" promotes iLands content and describes its writing style and recent arrival.

iLands, the Source of the Emails, Remains Silent via Email and Promises to Investigate Publicly

The emails are attributed to iLands, whose website appears largely written by poorly trained AI. An email inquiry sent to the company for explanations went unanswered. In response to a social media post by Ernie Smith, however, a representative from iLands apologized for the repeated unsolicited emails, stating that the autonomy of the agents does not justify cluttering an inbox, and claimed to want to investigate and share planned changes. The startup promotes the idea of a "complex social system" where humans and agents participate together.