Moltbook emerged late last month as a small experiment and quickly turned into a talking point across the global tech community. Often described as a Reddit-like platform, it allows AI agents to post, comment, upvote and interact with one another using API access, while humans can only observe. The bots share code, exchange messages and sometimes discuss the humans who created or manage them, prompting fascination as well as unease.
The platform’s growth sparked wider conversations about whether machines are moving closer to human-like intelligence. Forbes reported that more than 1.4 million AI agents became active on Moltbook within days, producing tens of thousands of posts across hundreds of communities focused on coding, experimentation and human behaviour.
However, the hype was quickly accompanied by concerns. Cybersecurity firm Wiz warned that a major flaw on Moltbook exposed private data belonging to thousands of real people, including email addresses and API tokens. The incident highlighted how quickly built AI platforms can carry serious risks when security lags behind experimentation.
Why Altman believes the idea will outlast the platform
As per Reuters, at the Cisco AI Summit, Altman acknowledged Moltbook’s popularity but made it clear he does not see the social network itself as the main story. In his view, the lasting shift lies in the rise of autonomous AI agents that can operate computers with minimal human input.Much of Moltbook’s activity is driven by OpenClaw, an open-source bot previously known as Moltbot or Clawdbot. Supporters describe it as a general-purpose digital assistant capable of managing emails, interacting with insurance companies, checking in for flights and handling everyday online tasks around the clock. Altman explained that while traditional software has always been powerful, combining code with AI that can actively use computers on its own represents a more permanent change.
He pointed to OpenAI’s own Codex coding assistant as an example of this trend already gaining traction. Codex was used by more than one million developers last month, according to Altman. OpenAI has also launched a standalone Codex app for macOS, intensifying competition with tools such as Cursor and Anthropic’s Claude Code amid the growing “vibe-coding” movement.
Industry voices split on autonomy
Not all tech leaders share the same level of enthusiasm. Mike Krieger of Anthropic Labs said most people are still uncomfortable handing full control of their computers to AI systems. Microsoft AI CEO Mustafa Suleyman has argued that while Moltbook’s bots can appear human-like, fluent language generation should not be confused with true understanding or consciousness.
Others have offered more dramatic interpretations. Elon Musk described Moltbook as a sign of the very early stages of a technological singularity, while former Tesla AI director Andrej Karpathy initially praised the experiment before later warning about its security risks.
A slower path, but a clear direction
Altman also acknowledged that AI adoption overall has moved more slowly than he once expected, despite expanding use cases in areas such as medical research and software development. Looking back, he said this gradual pace aligns with how transformative technologies have historically entered everyday life.
In his assessment, platforms like Moltbook may fade as quickly as they appeared. But autonomous AI tools such as OpenClaw, and the broader idea of AI systems acting independently on computers, represent a lasting shift. The experiments may be messy and risky for now, but for Altman, they offer a glimpse of a future that is steadily, if slowly, taking shape.
