What began as a weekend project named "WhatsApp Relay" has exploded into a massive open-source phenomenon, now officially rebranded as OpenClaw. Following legal nudges from Anthropic regarding the name "Clawd" and a brief stint as "Moltbot," creator Peter Steinberger has settled on a name that reflects the project's community-driven ethos and lobster-themed heritage. This platform distinguishes itself by allowing users to host their own AI assistants on local hardware—be it a laptop or a home server—rather than relying on centralized corporate servers.

The latest update isn't just a name change; it introduces robust support for new models like KIMI K2.5 and Xiaomi’s MiMo-V2-Flash, expanding the versatility of the assistant. By integrating directly with ubiquitous messaging apps like Telegram, Slack, and Discord, OpenClaw brings sophisticated automation to the platforms people already use daily. This "bring your own keys" approach ensures that personal data remains under the user's control, addressing growing privacy concerns in the age of generative AI.

Security remains a central pillar of the roadmap, with the team shipping 34 security-related commits to harden the codebase against vulnerabilities. While acknowledging that prompt injection—the act of tricking an AI into ignoring its original instructions—remains an industry-wide challenge, the project is moving toward machine-checkable security models. With a surge in community contributions and a plan to hire full-time maintainers, OpenClaw is transitioning from a solo hack into professional-grade infrastructure for decentralized AI agents.

What began as a weekend project named "WhatsApp Relay" has exploded into a massive open-source phenomenon, now officially rebranded as OpenClaw. Following legal nudges from Anthropic regarding the name "Clawd" and a brief stint as "Moltbot," creator Peter Steinberger has settled on a name that reflects the project's community-driven ethos and lobster-themed heritage. This platform distinguishes itself by allowing users to host their own AI assistants on local hardware—be it a laptop or a home server—rather than relying on centralized corporate servers.

The latest update isn't just a name change; it introduces robust support for new models like KIMI K2.5 and Xiaomi’s MiMo-V2-Flash, expanding the versatility of the assistant. By integrating directly with ubiquitous messaging apps like Telegram, Slack, and Discord, OpenClaw brings sophisticated automation to the platforms people already use daily. This "bring your own keys" approach ensures that personal data remains under the user's control, addressing growing privacy concerns in the age of generative AI.

Security remains a central pillar of the roadmap, with the team shipping 34 security-related commits to harden the codebase against vulnerabilities. While acknowledging that prompt injection—the act of tricking an AI into ignoring its original instructions—remains an industry-wide challenge, the project is moving toward machine-checkable security models. With a surge in community contributions and a plan to hire full-time maintainers, OpenClaw is transitioning from a solo hack into professional-grade infrastructure for decentralized AI agents.