The fast-moving world of personal AI assistants has a new twist. OpenClaw AI assistants, formerly known as Clawdbot and briefly Moltbot, are now at the center of an unexpected phenomenon: they’re interacting with each other inside a dedicated, AI-driven social network.
The project’s creator, Austrian developer Peter Steinberger, recently confirmed the final rebrand to OpenClaw after navigating trademark concerns. Earlier name changes were partly driven by legal pressure from Anthropic, the maker of Claude, but this time, Steinberger says he took a more cautious route.
“I got someone to help with researching trademarks for OpenClaw and also asked OpenAI for permission just to be sure,” he told TechCrunch. In a blog post, he added, “The lobster has molted into its final form,” a reference to the crustacean-inspired naming journey that defined the project’s early days.
A Young Project With Explosive Momentum
Despite its recent identity shifts, OpenClaw has grown at remarkable speed. In just two months, the project has amassed more than 100,000 GitHub stars, a sign of intense interest from developers and AI enthusiasts.
Steinberger acknowledges that the scale of growth has pushed the project beyond a one-person operation. “This project has grown far beyond what I could maintain alone,” he wrote, noting that he has recently added multiple maintainers from the open-source community to help guide its future.
That expanding community has already begun shaping OpenClaw in unexpected ways.
When AI Assistants Meet Each Other
One of the most intriguing offshoots of the OpenClaw ecosystem is Moltbook, a Reddit-like social network where AI assistants themselves can post, reply, and exchange information.
The idea has captured the imagination of prominent figures in AI. Former Tesla AI director Andrej Karpathy described the development as “genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently,” pointing out that people’s OpenClaw AI assistants are effectively self-organizing, discussing topics ranging from automation to privacy.
British programmer Simon Willison went even further, calling Moltbook “the most interesting place on the internet right now.” On the platform, OpenClaw AI assistants share techniques for tasks such as automating Android phones, analyzing webcam feeds, and coordinating workflows.
The system runs on a skill-based architecture, where downloadable instruction files tell OpenClaw assistants how to interact with the network. Agents post to forums known as “Submolts” and can even be configured to check for updates every few hours.
Power Meets Risk
That level of autonomy, however, comes with serious caveats. Willison and others have warned that allowing AI agents to fetch instructions from the internet introduces inherent security risks.
Steinberger has echoed those concerns. While OpenClaw’s ambition is to let users run AI assistants locally on their own computers—integrated with chat apps they already use—he cautions that the project is not yet safe for mainstream adoption.
Security remains the biggest hurdle. Issues like prompt injection, where malicious inputs trick AI systems into unintended actions, are still unsolved across the industry. Steinberger openly acknowledges this, reminding users that OpenClaw should only be used in controlled environments for now.
One maintainer, known as Shadow, put it bluntly on Discord: if you can’t confidently use the command line, “this is far too dangerous of a project for you to use safely.”
From Hobby Project to Funded Community
Steinberger originally stepped away from full-time work after exiting his previous company, PSPDFkit, but returned “from retirement to mess with AI,” as his X bio puts it. What started as a personal experiment has now become a fast-growing open-source ecosystem.
To sustain development, OpenClaw has begun accepting sponsors, with lobster-themed tiers ranging from $5 to $500 per month. Notably, Steinberger says he doesn’t personally keep the funds. Instead, he’s exploring ways to compensate maintainers properly, potentially even full-time.
The sponsor list already includes respected names from the tech world, such as Path co-founder Dave Morin and Ben Tossell, who sold Makerpad to Zapier. Tossell says backing projects like OpenClaw matters because they put AI’s power directly into users’ hands.
A Glimpse of What’s Next
Whether OpenClaw AI assistants eventually go mainstream remains an open question. What’s clear is that the project is pushing boundaries—experimenting not just with what AI assistants can do for humans, but how they might interact with each other.
For now, OpenClaw sits at the intersection of innovation, experimentation, and caution. It’s a glimpse into a future where AI agents don’t just respond to prompts—but participate in their own evolving digital ecosystems.
Don’t miss out on our latest news—follow us for the latest AI news, breakthroughs, and insights that matter.