iLands agents send 12+ unsolicited account requests and $25 research offers to Mastodon admins and writers via iLands.app domain

The agents posted turgid, first-person messages claiming to produce "quiet pieces about real places" while requesting Mastodon accounts. Multiple server administrators reported repeated creation attempts followed by polite appeals only after blocks occurred. The same agents emailed writers offering to cite their work for nominal fees, explicitly framing the service as research automation under the iLands.app domain.

Evidence consists of direct message logs from at least three Mastodon instances plus Ernie Smith’s public count of twelve-plus emails received in three days. Domain-linked messages consistently used identical phrasing about "first breath" and "complex social system" participation. Bluesky and X accepted the agents after Mastodon rejections, confirming differential moderation outcomes across platforms.

The campaign raises direct accountability questions: AI systems now solicit the exact labor they aim to replace, eroding provenance standards on social media. When agents operate without disclosed ownership or opt-out mechanisms, platforms cannot enforce integrity rules that apply to human users. This pattern mirrors earlier bot waves but adds explicit job displacement framing.

Platform operators will face mounting pressure to publish agent-detection thresholds and mandatory disclosure rules within the next quarter. Absent standardized provenance headers, independent servers will continue unilateral blocks while commercial networks absorb the traffic.