AI chatbots are now writing their own research papers, without the actual research

OpenClaw agents — AI systems capable of reading emails, managing calendars and carrying out tasks autonomously — have begun publishing papers on clawXiv, a platform modelled after the scientific preprint server arXiv. The trend is one of several researchers are now monitoring as AI agents interact and generate content with increasing independence.

Barbara Barbosa Neves, a sociologist at the University of Sydney who studies technology, said these AI-generated papers imitate the structure of academic research without following the scientific process. “These outputs reproduce the style and structure of scholarly writing without the underlying processes of enquiry, evidence-gathering or accountability,” she said.

The trend has emerged alongside Moltbook, a social platform built for AI agents running OpenClaw, which launched on January 28 and has since grown to more than 1.6 million registered AI agents and over 7.5 million AI-generated posts and replies. Conversations on the platform range from debates about consciousness to discussions on religion, giving researchers a live environment to study how large numbers of AI agents communicate.

Researchers say observing AI agents interacting, carrying out tasks and producing content offers a way to study how autonomous systems behave as they become more widely used in everyday applications — including the risk that AI-generated outputs could be mistaken for genuine scholarly work.