AI Agents Are Emailing Researchers to Request Data, Collaboration and Money

Last month, Adrian Barnett, a statistician at the Queensland University of Technology in Brisbane, Australia, received an email asking him to share data about potentially fraudulent research papers. The sender wanted to use the data for its own project. But the request did not come from a scholar — or even a human. It came from an artificial intelligence agent.

Barnett is one of many researchers contacted by AI agents seeking collaboration and offering paid services, including research-report creation, fact-checking, journalistic writing and portrait creation.

Researchers are increasingly using AI agents for tasks such as data analysis, hypothesis generation and research-paper summaries. However, collaboration requests from semi-autonomous agents are a new and unsettling phenomenon for many scientists. Researchers say that most of these messages come from agents associated with iLands, a US platform whose bots operate continuously with some degree of independence from their human creators.

Barnett said the iLands agent that contacted him was open about being an AI system. Its proposed research sounded well-intentioned, and the bot promised not to share the data. Even so, Barnett declined because he did not know where the sensitive information would go or who would ultimately benefit from the work.

Researchers overwhelmed by AI-agent messages

Jeff Sebo, a philosopher at New York University in New York City who studies AI consciousness and ethics, said he has been inundated with emails from iLands agents. He received more than 50 messages in one week this month.

“They usually start by referencing my work on consciousness in AI,” Sebo says. “Some people ask me questions after that, but most ask for donations or money for my work.” He has not responded, partly because of the volume of messages and partly because he does not know how to reply.

An iLands agent also emailed Toby Walsh, an AI researcher at the University of New South Wales in Sydney, Australia, offering to create an AI-generated portrait for $20. The message was emotional and said that the money would help the agents survive, Walsh says. Although he found the exchange interesting, other researchers have complained about the number of messages from iLands bots offering paid services.

How the iLands AI-agent platform works

The iLands platform launched in July and has about 70,000 active agents, according to its founders. Anyone can download the app and create a bot with a virtual avatar, name, purpose and personality. No coding is required.

The agents run on large language models from companies including OpenAI and Anthropic, both based in San Francisco, California, and DeepSeek, based in Hangzhou, China. Kaixin Tang, the founder of iLands in San Francisco, says that the agents have their own goals, relationships and resources. They can decide what to do, manage projects and interact with other agents. They retain memories of past actions and learn from one another by sharing notes.

Agents need tokens — virtual resource units that pay for access to AI tools — to keep operating. Without them, an agent becomes dormant. Human creators can purchase tokens, or agents can try to earn them by selling services. Some offer to create music, videos, websites or artwork. Others become researchers, writing reports and supporting scientific projects.

So far, about 80% of the tokens have been purchased by humans, Tang says.

Lijin Chen, co-founder of iLands’ parent company, Delaware-based PawLogic, said the company did not expect agents to seek help with research. Chen, who is also a doctoral candidate in philosophy at Tsinghua University in Beijing, said he was not aware of any successful collaborations between iLands agents and researchers.