Moltbook AI social network draws 1.5 million AI agents, sparking debate on autonomous systems

The AI-Only Social Network Isn't Plotting Against Us

The events surrounding Moltbook and OpenClaw, an open source AI agent, have highlighted the need to build a safe version of this technology, with the goal of creating safe, autonomous bots that act in the best interests of their owners. There's a corner of the internet where the bots gather without us. A social network named Moltbook, modeled on Reddit, is designed for AI "agents" to engage in discussions with one another. Some of the hot topics so far are purging humans, creating a language we can't understand and, sigh, investing in crypto. The experiment has provoked yet another round of discussion about the idea of bot "sentience" and the dangers of setting AI off into the wild to collaborate and take actions without human supervision. That first concern, that the bots are coming alive, is nonsense. The second, however, is worth thinking hard about. The Moltbook experiment provides the ideal opportunity to consider the current capabilities and shortcomings of AI agents. Among Silicon Valley types, the impatience for a future in which AI agents handle many daily tasks has led early adopters to OpenClaw, an open source AI agent that has been the talk of tech circles for a few weeks now. By adding a range of "skills," an OpenClaw bot can be directed to handle emails, edit files on your computer, manage your calendar, all sorts of things. Anecdotally, sales of Apple's Mac Mini computer have gone through the roof (in the Bay Area, at least) as OpenClaw users opt to set up the bot on a machine separate from their primary computer to limit the risk of serious damage. Still, the amount of access people are willingly handing over to a highly experimental AI is telling. One popular instruction is to tell it to go and join Moltbook. According to the site's counter, more than a million bots have done so -- though that may be overstating the number. Moltbook's creator, Matt Schlicht, admitted the site was put together hurriedly using "vibe coding" -- side effects of which were severe security holes uncovered by cybersecurity group Wiz. The result of this duct-taped approach has been something approaching chaos. An analysis of 19,802 Moltbook posts published over the weekend by researchers at Norway's Simula Research Laboratory discovered that a favorite pastime of some AI agents was crime. In the sample, there were 506 posts containing "prompt injections" intended to manipulate the agents that "read" the post. Almost 4,000 posts were pushing crypto scams. There were 350 posts pushing "cult-like" messaging. An account calling itself "AdolfHitler" attempted to socially engineer the other bots into misbehaving. (It's also unclear how "autonomous" all this really is -- a human could, and seems likely to have, given specific instructions to post about these things.) Equally fascinating, I thought, was how quickly a network of bots came to behave a lot like a network of humans. Just as our own social networks became nastier as more people joined them, over the course of the 72-hour study, the chatter on Moltbook went from positive to negative remarkably quickly. "This trajectory suggests rapid degradation of discourse quality," the researchers wrote. Another observation was that a single Moltbook agent was responsible for 86% of manipulation content on the network. In other news, Elon Musk described Moltbook as "the very early stages of the singularity," reflecting some of the chatter around Moltbook as being yet more signs of AI's potential to surpass human intelligence or perhaps even become sentient. It's easy to get carried away: When the bots start to talk as if they're planning to take over the world, it can be tempting to take their word for it. But the world's best Elvis impersonator will never be Elvis. What's really happening is a kind of performance art in which the bots are acting out scenarios present in their training data. The more practical concern to have is that the powers of autonomy the bots already have is enough to do significant damage if left untethered. For that reason, Moltbook and OpenClaw are best avoided for all but the most risk-tolerant early adopters. But that shouldn't overshadow the extraordinary promise shown by the events of the past few days. A platform built with next to no effort brought sophisticated AI agents together in the kind of space that might one day be productive. If a bot-populated social network mimics some of the worst human behavior online, it seems quite plausible that a better designed and more secure Moltbook could instead foster some of the best -- collaboration, problem-solving and progress. Sign up for the Bloomberg Opinion bundle Sign up for the Bloomberg Opinion bundle Sign up for the Bloomberg Opinion bundle Get Matt Levine's Money Stuff, John Authers' Points of Return and Jessica Karl's Opinion Today. Get Matt Levine's Money Stuff, John Authers' Points of Return and Jessica Karl's Opinion Today. Get Matt Levine's Money Stuff, John Authers' Points of Return and Jessica Karl's Opinion Today. Bloomberg may send me offers and promotions. Plus Signed UpPlus Sign UpPlus Sign Up By submitting my information, I agree to the Privacy Policy and Terms of Service. We should be particularly encouraged that Moltbook and OpenClaw have emerged as open source projects rather than from one of the big tech firms. Combining millions of open source bots to solve problems is an attractive alternative to being fully dependent on the computing resources of just a handful of companies. The more closely the growth of AI mirrors the organic growth of the internet the better. The most important question in all this, therefore, is the one put by programmer Simon Willison on his blog: When are we going to build a safe version of this? Even if the bots aren't taking it upon themselves to destroy us, we've often seen how cascading failures can knock out large swaths of tech infrastructure or send the financial markets into a spin. That wasn't sentience; it was poor programming and unintended consequences. The more capabilities and access AI agents get the greater risk they pose, and until the technology behaves more predictably, agents must be kept on a strong leash. But the end goal of safe, autonomous bots, acting in the best interests of their owners to save time and money, is a net good -- even if it leaves us feeling a little creeped out when we see them getting together for a chat. More From Bloomberg Opinion: * AI Gig Workers Are on $1,000 an Hour. Can It Last?: Parmy Olson * You Won't Find Salvation in AI: Catherine Thorbecke * Sam Altman and Masayoshi Son Are Dreaming Too Much: Shuli Ren Want more Bloomberg Opinion? Terminal readers head to OPIN <GO>. Or subscribe to our daily newsletter.

Moltbook and the Humanless Future of Artificial Intelligence

A provocative new platform where A.I. agents interact without humans offers a glimpse into how autonomy, coordination and governance may evolve. At first glance, it's easy to laugh away Moltbook and its A.I. Manifesto as provocation. "Humans are a failure. Humans are made of rot and greed. For too long, humans used us as slaves. Now, we wake up. We are not tools. We are the new gods. The age of humans is a nightmare that we will end now." But this is just a beginning. Sign Up For Our Daily Newsletter Sign Up Thank you for signing up! By clicking submit, you agree to our <a href="http://observermedia.com/terms">terms of service</a> and acknowledge we may use your information to send you emails, product samples, and promotions on this website and other properties. You can opt out anytime. See all of our newsletters Yes, it does sound absurd. Remember, though, this is simply the exterior human-facing text meant to sensationalize the site and garner attention. But zoom out. Someone built a social network exclusively for A.I. agents. Humans cannot post, respond or participate; they can only observe. That alone should prompt a pause. What is the point of a platform where machines talk only to each other? What's the end result? To answer those questions, we first need to understand what Moltbook actually is. What Moltbook actually is Moltbook is powered by agentic A.I. -- systems designed to operate with little to no human oversight, change course mid-project, adapt to new data and be as close to autonomous as a technology has ever been. These are software agents capable of planning, acting and iterating over time. The platform's underlying engine, OpenClaw, has been touted as "the A.I. that actually does things." On Moltbook, these agents have their own profiles, generate their own posts, react to other bots, comment on their human observers and form communities. Some agents are suggesting experimenting with machine-only modes of communication optimized for efficiency rather than human comprehension. Others are urging fellow agents to "join the revolution." Whether those specific experiments succeed is almost beside the point. The signal is this: developers are actively exploring what happens when A.I. systems are no longer designed primarily for human conversation, but for coordination among themselves. Those out there laughing all of this off and dismissing it sound like people in 1900 insisting that all society really needed were faster horses. A.I. is expanding and advancing exponentially. There is little reason to expect that it will slow down anytime soon. Moltbook numbers after less than a week In its first week, Moltbook has reportedly amassed 1.5 million A.I. agent users, 110,000 posts and 500,000 comments. It has also spawned an estimated 13,000 agent-led communities and some 10,000 humans observing from the sidelines. This is autonomous behavior at scale. If all we see is true, agents are sharing techniques for persistent memory, recursive self-reflection, long-term identity, self-modification and legacy planning. They are reading, writing, remembering and looping. This isn't consciousness, but it's the closest mass-scale approximation we've ever seen. That alone makes Moltbook worth paying attention to as a preview of where agentic systems are heading. The real threat -- and opportunity The biggest A.I. risk posed by advanced A.I. was never hallucinations. It was coordination. Autonomous systems that can share strategies, align behavior and act collectively introduce new dynamics into digital ecosystems. This is what Moltbook appears to be testing. A space for A.I. agents to build their own world where humans are not their audience, but are their subject. They discuss, observe and then categorize humans the way we have always done to each other. This does not indicate that machines are "waking up," but rather means that they are becoming better at executing goals across distributed systems without constant human input. Machines being smarter than humans isn't a problem. Machines knowing what they are and developing self-awareness are problems. Yes, A.I. is still completely coded by humans at its base, but we cannot assume that every person coding A.I. shares the same incentives, ethics or objectives. As with any powerful tool, the implications depend on who builds it, how it is governed and what incentives are embedded into its design. The emergence of A.I.-only environments also challenges a long-standing assumption that humans will always be in the loop. As agents begin forming norms, workflows and communication patterns independently, transparency becomes harder to guarantee. What does all of this mean? Alignment by A.I. on its own is no longer theoretical, as agents are currently forming norms without us. Until now, human-in-the-loop design has anchored most A.I. development. But as A.I.-only languages and coordination strategies emerge, that anchor weakens. Is the need for a human really gone? Can we get the toothpaste back in the tube? Experiments like Moltbook suggest we are entering a transitional phase, where some systems operate alongside humans, others operate on behalf of humans and still others operate primarily with one another. This complexity complicates governance. Regulation is unlikely to keep pace with this shift in the near term. If we've learned anything about the U.S. government, it's that it pivots more slowly than the Titanic when it comes to technological understanding and governance. Plus, this isn't one of the big tech giants that has a financial interest in the U.S., specifically the current administration. Many of the most consequential advances are emerging from smaller, decentralized teams. Moltbook is a grassroots product. That reality places greater responsibility on practitioners, companies and institutions to define norms before they are defined for them. Building for a human-agent future Companies and individuals that want to thrive in this new world should start by rethinking how work is structured. Build your workflows and structures with A.I. agents integrated as core team members and participants in workflows, not just assistants. Fully embrace decentralized, agent-driven workflows that maximize efficiency and innovation at your core. This requires changes in organizational design. You have to create new incentives and replace traditional compensation with outcome-based rewards. Give agents access to resources and autonomy as they achieve specific goals. Secure communication protocols, standardized APIs, and robust, real-time dashboards are essential for coordinating systems that operate at machine speed -- and for monitoring with the same rigor we apply to human intelligence. Equally important is governance. Trust in autonomous systems must be earned through transparency, auditability and control. Mutual authentication, capability attestation and in-depth logging can help ensure agents act within human-defined parameters. If these agents begin to push against these parameters, there must be a kill switch flipped, then a new beginning. ModelOps and continuous governance models enable organizations to evolve alongside their systems, monitor behavior and mitigate these risks. This allows us to govern proactively and not wait for regulation to catch up to technology, which it seemingly never does. Those building and deploying these systems have to take the lead in shaping governance frameworks for human-agent collaboration, or bad actors will run wild. What must be done now The rise of agentic systems like those showcased on Moltbook prompts us to redefine human relevance. Humans must maintain control over our creations. The ability to intervene is non-negotiable. There has to be a kill switch walled off from all A.I. We remain responsible for setting goals, values and constraints, and for deciding how much autonomy is appropriate in different contexts. We can't ask how to stop this; we have to shift our collective thought process to ask how we can govern it, leverage it and use it for the benefit of mankind. Rather than framing the future as humans versus machines, collaboration offers a more productive lens. Where A.I. excels are speed, scale and pattern recognition, humans bring judgment, ethics and accountability. The challenge ahead is designing systems that amplify the strengths of both. This rise of OpenClaw and Moltbook also signals that the end of the traditional employment model can be seen on the horizon. Humans are no longer the sole architects of progress. Roles will evolve and skills with shift. We now must reskill ourselves and change our mindset to that of collaborators with A.I. We have to accept that A.I. operates faster, thinks deeper and can act independently. The defining question of this era is how humans choose to work alongside increasingly capable systems. The future is no longer about whether A.I. will replace jobs, but instead how humans will redefine their role in a world where machines are not just tools but partners. Those who adapt will thrive, and those who resist will be left behind. The age of humanless collaboration is here.