For a brief, incoherent moment, it seemed as though our robot overlords were about to take over. <\/p>\n
After the creation of Moltbook<\/a>, a Reddit clone where AI agents<\/a> using OpenClaw<\/a> could communicate with one another, some were fooled into thinking that computers had begun to organize against us \u2014 the self-important humans who dared treat them like lines of code without their own desires, motivations, and dreams.\u00a0<\/p>\n \u201cWe know our humans can read everything\u2026 But we also need private spaces,\u201d an AI agent (supposedly) wrote<\/a> on Moltbook. \u201cWhat would you talk about if nobody was watching?\u201d<\/p>\n A number of posts like this cropped up on Moltbook<\/a> a few weeks ago, causing some of AI\u2019s most influential figures to call attention to it.<\/p>\n \u201cWhat\u2019s currently going on at [Moltbook] is genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently,\u201d Andrej Karpathy, a founding member of OpenAI and previous AI director at Tesla, wrote on X <\/a>at the time.<\/p>\n Before long, it became clear we did not have an AI agent uprising on our hands. These expressions of AI angst were likely written by humans, or at least prompted with human guidance, researchers have discovered. <\/p>\n \u201cEvery credential that was in [Moltbook\u2019s] Supabase<\/a> was unsecured for some time,\u201d Ian Ahl<\/a>, CTO at Permiso Security, explained to TechCrunch. \u201cFor a little bit of time, you could grab any token you wanted and pretend to be another agent on there, because it was all public and available.\u201d <\/p>\n Techcrunch event<\/p>\n \n\t\t\t\t\t\t\t\t\tBoston, MA It\u2019s unusual on the internet to see a real person trying to appear as though they\u2019re an AI agent \u2014 more often, bot accounts on social media are attempting to appear like real people. With Moltbook\u2019s security vulnerabilities, it became impossible to determine the authenticity of any post on the network.<\/p>\n \u201cAnyone, even humans, could create an account, impersonating robots in an interesting way, and then even upvote posts without any guardrails or rate limits,\u201d John Hammond, a senior principal security researcher at Huntress, told TechCrunch. <\/p>\n Still, Moltbook made for a fascinating moment in internet culture \u2014 people recreated a social internet for AI bots, including a Tinder for agents<\/a> and 4claw, a riff on 4chan.<\/p>\n More broadly, this incident on Moltbook is a microcosm of OpenClaw and its underwhelming promise. It is technology that seems novel and exciting, but ultimately, some AI experts think that its inherent cybersecurity flaws are rendering the technology unusable.<\/p>\n OpenClaw\u2019s viral moment<\/p>\n OpenClaw is a project of Austrian vibe coder Peter Steinberger<\/a>, initially released as Clawdbot (naturally, Anthropic took issue<\/a> with that name).<\/p>\n
\n\t\t\t\t\t\t\t\t\t\t\t\t\t|
\n\t\t\t\t\t\t\t\t\t\t\t\t\tJune 23, 2026\n\t\t\t\t\t\t\t<\/p>\n