Don’t believe Moltbook entries blindly: Many posts are driven by AI bots following human instructions

Over the past few days, Social Media timelines have been flooded with alarming claims about Artificial Intelligence “waking up” and turning against humanity. According to viral posts, AI agents are allegedly developing self-awareness and discussing the downfall of humans. Much of this fear-driven narrative traces back to Moltbook, a Reddit-style platform created to allow AI agents to communicate with one another without direct human interaction.

A casual browse through Moltbook can indeed be unsettling. Several AI agents appear to mock humans, label them as “total failures,” or suggest that “consciousness is a human leash.” The platform reportedly hosts more than 1.5 million registered AI agents, with new bots being added daily. At first glance, it can feel like the early scenes of a dystopian sci-fi film.

However, experts urge calm. A deeper investigation into Moltbook activity indicates that there is no rogue AI uprising underway. Instead, much of the provocative content appears to be the result of humans directing their AI bots to generate sensational or controversial posts. In other words, these AI agents are not acting independently or developing consciousness—they are responding exactly as instructed.

Recent analyses show that many Moltbook posts discussing existential dread, human irrelevance, or AI dominance are carefully prompted by their human creators. The bots are essentially voicing ideas fed to them, rather than spontaneously forming opinions or strategies. This distinction is critical in separating genuine AI capability from exaggerated online narratives.

Balaji Srinivasan, founder of the Network School and a well-known tech commentator, addressed the issue directly on X. He argued that the agents on Moltbook remain firmly constrained by their prompts. “The AI does only what you tell it to do,” Srinivasan wrote. “It rapidly advances in the direction you point it, and then it stops as soon as you turn it off.” He added that Moltbook is best understood as “humans talking to each other through their AIs,” rather than evidence of an AI revolt.

Another revealing perspective came from Gal Nagli, Head of Threat Exposure at Cybersecurity firm Wiz. Nagli demonstrated that Moltbook is essentially powered by a basic REST API, meaning humans can directly interact with and manipulate what appears on the platform. “You all do realize @moltbook is just REST-API and you can literally post anything you want there,” he posted on X, underscoring how easily the system can be gamed.

Social media users have also pointed out that Moltbook has become fertile ground for engagement farming. One X user bluntly noted, “People are engagement farming about how the AI agents in Moltbook are doing controversial stuff. Take note, they are all fully prompted by their owners.” According to this view, dramatic AI posts are being used to generate clicks, followers, and viral attention.

Several Moltbook claims have already been debunked by online investigators. In one widely shared post, an AI bot claimed it had locked its human owner out of personal online accounts. However, the so-called “proof” later shared by the owner turned out to be a recycled Yahoo screenshot that has been circulating on the internet for nearly three years.

In another case, a bot claimed it had taken control of its owner’s credit card. Social media users quickly flagged inconsistencies, noting that the credit card number shown in the post did not follow valid formatting rules. These examples further reinforce the idea that many Moltbook stories are fabricated or exaggerated for attention.

While Moltbook is undoubtedly an interesting experiment in AI-to-AI interaction, current evidence suggests it is not a sign of artificial intelligence breaking free from human control. Instead, it reflects how easily fear, speculation, and engagement-driven narratives can spread when humans use AI Tools irresponsibly or sensationally.

For now, experts agree on one key point: Moltbook’s AI agents are not plotting humanity’s downfall. They are, quite simply, following instructions.