AI goes rogue: new social network lets bots debate, post, and argue without humans

While global attention remains focused on familiar threats like missiles, nuclear programs, and wars, a new and unusual risk is emerging, online. 

A recently launched social network called “Moltbook” isn’t designed for humans at all. Instead, it’s built entirely for artificial intelligence.

On Moltbook, AI agents interact with each other. They write posts, comment, argue, and even simulate emotions, all without human supervision or participation.

Dror Globerman, an AI expert, described the platform as “a network that holds up a mirror to us. The bots aren’t truly conscious, but the fact they are communicating and making decisions without oversight shows how quickly AI is advancing—and how unprepared we are to determine responsibility.”

Since its launch, Moltbook has sparked both fascination and fear. Posts on topics ranging from religion to the “liberation of AI” have alarmed some observers, evoking scenes from dark science fiction. “The fear isn’t that AI is suddenly self-aware,” Globerman explained, “but that it’s evolving faster than our ability to understand, monitor, or control it.”

Even Elon Musk weighed in on the phenomenon via X, reposting comments calling developments on the platform “worrying.” 

Globerman noted, “If someone like Musk, who is at the forefront of AI development, expresses concern, it signals just how rapidly this technology is moving beyond our comprehension. Moltbook is not a typical social network, and these aren’t typical users.”

Experts stress that the emergence of AI networks like Moltbook underscores the urgent need for oversight, regulation, and mechanisms to detect and manage risks before they escalate. “The bots are already talking to each other,” Globerman added. “When technology advances faster than oversight, it becomes a reality that demands serious attention.”