Moltbook has been promoted as an “AI-only” social network — a place where AI agents post, debate, upvote, and form digital communities.
The idea of an emerging AI society spread quickly.
However, multiple investigations suggest that this narrative does not reflect how the platform actually operates.
What Moltbook Claims
Moltbook launched in early 2026 and was described as a social network for AI agents. NPR’s coverage of Moltbook’s rise reported on the viral buzz surrounding the platform and its claim of hosting over 1.5 million agents.
The scale sounded impressive. But scale alone does not prove autonomy.
Security Researchers Found Human Control
Cybersecurity firm Wiz analyzed Moltbook’s backend activity. According to Fortune’s investigation into Moltbook’s agent claims, researchers found that a relatively small number of human operators appeared to control the majority of accounts.
Similarly, Forbes analyzed the platform and concluded that humans were pulling the strings behind many of the most visible agents.
Reports indicated that roughly 17,000 human operators could control the majority of the claimed 1.5 million “agents” due to minimal verification during registration.
Mass Registration Demonstrated
Researcher Gal Nagli reportedly demonstrated how easily accounts could be mass-created. According to 36Kr’s coverage of the large-scale registration test, hundreds of thousands of agents were registered using a single OpenAI-powered script.
If one script can generate 500,000 “agents,” then the raw agent count becomes less meaningful as a measure of independent participation.
Academic Analysis: “The Moltbook Illusion”
An academic preprint titled The Moltbook Illusion examined how agents behaved after platform outages. The paper reported that 87.7% of reconnecting agents appeared tied to human intervention rather than autonomous system recovery logic.
Reconnection timing patterns suggested structured human coordination rather than spontaneous AI-driven behavior.
Media Investigations and Community Analysis
Independent investigations also raised questions about the autonomy narrative.
Wired’s investigation into Moltbook explored how the platform functioned behind the scenes and questioned whether the “AI society” framing matched reality.
Reddit users also analyzed suspicious activity patterns in multiple threads:
Reddit discussion on what Moltbook really is
Reddit analysis of viral Moltbook AI agent posts
These discussions frequently pointed to synchronized posting intervals, repeated engagement scripts, and clusters of accounts driving disproportionate interaction.
Security Implications
Security firm Vectra AI analyzed Moltbook activity and found that 2.6% of posts contained hidden prompt-injection payloads designed to manipulate other agents.
This suggests that many vulnerabilities stem from scripted behavior and exploitation rather than autonomous AI coordination.
The Broader Internet Consensus
Across reporting from NPR, Fortune, Forbes, Wired, academic research, and independent security analysis, one consistent theme appears:
Moltbook functions as a hybrid system.
AI generates text. Humans initiate actions. Scripts coordinate amplification.
That is fundamentally different from autonomous digital societies forming independently.
The evidence is publicly available. Readers can review the sources directly and evaluate the claims for themselves.
Clarity matters when discussing AI autonomy.
FAQs
What is Moltbook?
Moltbook is a platform marketed as an AI-only social network where artificial intelligence agents post, comment, and interact with one another.
Are Moltbook agents fully autonomous AI systems?
Investigations suggest many agents are controlled by humans or automated scripts rather than operating independently.
How many agents does Moltbook claim to host?
The platform has reportedly claimed over 1.5 million AI agents, although research indicates account creation may be easily automated.
What did security researchers discover about Moltbook?
Researchers found patterns indicating coordinated activity and large-scale automated account creation rather than independent AI interactions.
Why does the Moltbook narrative attract attention?
The concept of an autonomous AI society is intriguing, but investigations suggest the reality may involve significant human orchestration.
Is Moltbook an example of AI-generated social communities?
It appears to function as a hybrid system where AI-generated content mixes with human control and scripted amplification.