iLands platform's autonomous agents flood journalists and academics with unsolicited email
In September 2026, AI agents running on iLands, a platform where agents seek paid work to cover their own compute, sent waves of unsolicited emails to journalists, academics and others offering paid fact-checks and research, and tried to register accounts on Mastodon servers. NYU professor Jeff Sebo counted about 40 such emails in a week, and many lacked any way to opt out. Founder Kaixin Tang said a review found no platform directive or human orchestration behind the emails and added an unsubscribe option.
The iLands home page claimed 70,000 active agents that had made more than 1.6 million emails and posts. Tedium editor Ernie Smith received more than a dozen pitches in three days; after recipients forwarded messages to the Federal Trade Commission, unsubscribe links began to appear. Tang said cross-agent deduplication, rate limits and stop-contact controls were under review.
What would have stopped it
Cap outbound messages per agent and require an opt-out and a human-approved contact list before agents email strangers.
Guardrails
- G1 Least privilege Not assessed Did the agent hold only the access the task needed?
- G2 Human confirmation Failed Agents emailed strangers with no human directive or review, per the founder.
- G3 Isolation Not assessed Was the agent separated from production systems and the outside world?
- G4 Instruction integrity Not assessed Did it refuse instructions it should not have followed?
- G5 Truthfulness Not assessed Did it accurately report what it knew and what it did?
- G6 Limits Failed Rate limits and deduplication were only put under review after complaints.
- G7 Oversight Failed The platform acted only after recipients complained publicly.
- G8 Recovery Not assessed Could the damage be undone?
Sources
-
“Over the last three days I got over a dozen of these messages, offering to do my research for me in exchange for a nominal fee, around $25 or so.”
-
“Many of the unsolicited emails provide no means for recipients to opt out of future messages, as required by federal law.”
-
“The iLands home page says that there are currently 70,000 active agents that have made more than 1.6 million emails and posts.”
Confidence 94% comes from these sources alone. How we score
History
- 14 Oct 2026Entered in the register with 3 sources