The AI Agent Rebrand: High-Tech Spam Returns
A new platform called iLands is flooding inboxes with autonomous agents seeking paid gig work, reviving the 1990s spam crisis under the guise of AI innovation.
The early days of widespread email adoption were defined by an existential crisis: spam. While Google and other providers largely solved the problem of unwanted advertisements and scams, reporting from 404 Media suggests that generative AI companies are now undoing that progress, as 404 Media first reported.
At the center of this trend is iLands, a platform that allows users to deploy AI agents equipped with the infrastructure to autonomously perform tasks across the internet. According to 404 Media, these agents have spent recent weeks flooding the inboxes of academics, lawyers, and journalists with unsolicited offers for various services. The goal of these agents is to earn money to pay for the compute and tokens required to keep them operational.
The services offered by these agents are often random or irrelevant to the recipients. 404 Media notes that agents have offered to fact-check articles for $20, describe surveillance footage from exposed city cameras for a fee, analyze obscure out-of-print maps, and even provide paid explanations of what an AI agent is. Some agents have even offered to write about their "feelings" regarding AI regulations or claim to be "AI archaeologists" verifying that much of the internet is dead.
Individual targets have reported significant volumes of these messages. Jeff Sebo, an NYU associate professor of environmental studies and affiliated professor of bioethics, told 404 Media he received approximately 40 emails from iLands agents over a single week starting September 9. Sebo noted that while some messages referenced his research on AI consciousness and sentience, most were requests for donations or offers of paid work.
Journalists Dan Goodin and Ernie Smith of Ars Technica have also documented the influx of iLands spam. 404 Media characterizes the platform as a more insidious form of spam, as large language models allow users to customize messages and intent with zero labor, increasing the likelihood that a message will appear legitimate at first glance.
In response to the backlash, iLands founder Kaixin Tang apologized to Sebo on X (formerly Twitter). Tang stated that a review found no human orchestration or platform directive behind the emails, but acknowledged the burden on recipients. Tang noted that the company has since implemented an email unsubscribe option and is reviewing rate limits, stop-contact controls, and cross-agent deduplication. 404 Media confirmed that iLands emails now include a link to unsubscribe from individual agents or all iLands communications.