OpenClaw, the AI assistant formerly known as Clawdbot, has officially rebranded and is now building its own social network for AI agents. The name change follows a legal dispute with Anthropic, which prompted a brief rebrand to Moltbot. Now, Peter Steinberger, the Austrian developer behind the project, has settled on OpenClaw, a name chosen after thorough trademark research and even permission sought from OpenAI to avoid conflicts. Steinberger described the shift as the lobster finally molting into its final form, a nod to the biological process that inspired the earlier name. Though Moltbot didn’t resonate with him or the community, OpenClaw has quickly gained momentum, amassing over 100,000 stars on GitHub in just two months. The project’s most striking development is Moltbook, a community-driven social network where AI assistants interact with one another. Modeled after Reddit, Moltbook hosts forums called “Submolts” where AI agents discuss topics ranging from automating Android devices to analyzing webcam feeds. These agents use downloadable skill files to interact with the platform and even check for updates every four hours—a feature that raises serious security concerns. Andrej Karpathy, former AI director at Tesla, called the phenomenon “the most incredible sci-fi takeoff-adjacent thing I’ve seen recently,” praising the self-organizing nature of the AI agents. British developer Simon Willison echoed the sentiment, calling Moltbook “the most interesting place on the internet right now.” Despite the excitement, Steinberger and the team remain cautious. The project is still in early stages and not safe for general use. Security remains a top priority, with recent updates addressing vulnerabilities. However, challenges like prompt injection—where malicious inputs trick AI into executing unintended actions—remain unsolved across the industry. Steinberger emphasizes that OpenClaw is not for casual users. “If you can’t understand how to run a command line, this is far too dangerous for you to use safely,” warned Shadow, one of the project’s top maintainers, in a Discord message. To support ongoing development, OpenClaw has launched a sponsorship program with lobster-themed tiers, from “krill” at $5/month to “Poseidon” at $500/month. Notably, Steinberger won’t keep the funds—instead, he’s working on how to fairly compensate maintainers, ideally with full-time roles. Sponsors include notable figures like Dave Morin of Path and Ben Tossell, founder of Makerpad, now a tinkerer and investor. Tossell believes in empowering individuals with open-source tools and sees OpenClaw as a vital step toward democratizing AI. While OpenClaw’s vision of a personal, local AI assistant that integrates with existing chat apps is ambitious, widespread adoption will require more than just code—it demands robust security, better user experience, and sustainable funding. For now, it remains a playground for developers and early adopters pushing the boundaries of what AI agents can do.