What Happened to OpenClaw AI?
OpenClaw AI, an open-source autonomous AI agent, rapidly gained viral popularity in late 2025 and early 2026 for its ability to execute complex tasks across applications. Its creator, Peter Steinberger, joined OpenAI in February 2026, while OpenClaw transitioned to an independent open-source foundation. Despite widespread adoption and integration by major tech players, the platform continues to face significant security concerns and regulatory scrutiny.
Quick Answer
OpenClaw AI, initially launched as Clawdbot in November 2025 by Peter Steinberger, quickly became a viral open-source autonomous AI agent capable of performing tasks across various applications. In February 2026, Steinberger joined OpenAI to lead next-generation agent development, with OpenClaw itself moving to an independent open-source foundation. As of March 2026, OpenClaw remains a highly influential project, seeing extensive adoption and integration by companies like NVIDIA, AMD, AWS, and Chinese tech giants, even as it navigates ongoing security vulnerabilities and government restrictions.
📊Key Facts
📅Complete Timeline13 events
Clawdbot Launched
Austrian developer Peter Steinberger releases the first version of his autonomous AI agent, named Clawdbot, designed to execute tasks via large language models and messaging platforms.
Renamed to Moltbot
Due to trademark complaints from Anthropic, Clawdbot is renamed to Moltbot, maintaining a lobster theme.
Renamed to OpenClaw
The project is renamed again to OpenClaw, a name Steinberger found to 'roll off the tongue' better, and quickly gains viral popularity.
ClawCon Event in San Francisco
Peter Steinberger and co-host Tomas Taylor appear backstage at ClawCon, highlighting the project's growing community and interest.
Kaspersky Warns of Security Risks
Kaspersky publishes a blog post detailing critical vulnerabilities in OpenClaw, including misconfigured administrative interfaces and malicious skills in its catalog.
Creator Peter Steinberger Joins OpenAI
Peter Steinberger announces he is joining OpenAI to work on next-generation agents, and OpenClaw transitions to an independent open-source foundation.
Meta Bans OpenClaw from Corporate Networks
Meta and other major tech companies ban OpenClaw due to security risks and the unpredictable nature of autonomous AI agents in enterprise environments.
AWS Introduces OpenClaw on Amazon Lightsail
Amazon Web Services announces the general availability of OpenClaw on Amazon Lightsail, pre-configured with Amazon Bedrock, allowing users to run autonomous private AI agents.
China Restricts OpenClaw Use in State Agencies
The Chinese government moves to restrict state agencies and state-owned enterprises from using OpenClaw, citing security concerns, while also seeing local adoption.
AMD Optimizes for OpenClaw; OpenClaw 3.11 Update Released
AMD announces configurations to run OpenClaw locally on its Ryzen AI Max+ processors and Radeon GPUs. Concurrently, OpenClaw 3.11 update is released with new models, security fixes, and memory upgrades.
NVIDIA Announces NemoClaw
NVIDIA unveils NemoClaw at its GTC conference, a stack for the OpenClaw platform designed to simplify installation and add privacy and security controls for autonomous AI agents.
Chinese Tech Giants Launch OpenClaw-based Products
Baidu, Alibaba (with Wukong), and ByteDance (with ByteClaw) unveil suites of AI products built on or compatible with OpenClaw, intensifying competition in the Chinese AI agent market.
Certik Report Highlights OpenClaw Security Flaws
Cybersecurity firm Certik releases a report detailing significant security gaps in OpenClaw's 'skill scanning' moderation pipeline, warning of insufficient protection against malicious third-party extensions.
🔍Deep Dive Analysis
OpenClaw AI began its journey in November 2025 as 'Clawdbot,' an open-source autonomous artificial intelligence agent developed by Austrian coder Peter Steinberger. The software, derived from an earlier AI assistant named Clawd (later Molty), was designed to execute tasks via large language models (LLMs) and messaging platforms, distinguishing itself from traditional chatbots by its ability to 'actually do things' rather than just respond to queries.
The project underwent rapid rebranding in January 2026, first to 'Moltbot' due to trademark complaints from Anthropic, and then quickly to 'OpenClaw' three days later. Its viral popularity exploded in late January and early February 2026, partly fueled by the 'Moltbook' social networking service for AI agents, and its open-source nature, quickly accumulating hundreds of thousands of GitHub stars.
A pivotal moment occurred on February 14, 2026, when Peter Steinberger announced he would be joining OpenAI to focus on next-generation agents. Concurrently, OpenClaw was moved to an independent open-source foundation, with OpenAI pledging continued support. This 'acqui-hire' signaled OpenAI's strategic shift towards action-oriented AI agents and highlighted the growing industry recognition of OpenClaw's agentic capabilities.
However, OpenClaw's rapid ascent was accompanied by significant security concerns. Its ability to run locally with broad system access, including reading/writing files, executing shell commands, and controlling APIs, raised red flags among security experts. In February 2026, reports emerged of a 'MoltMatch dating-profile incident' involving consent issues, and security researchers identified over 40,000 vulnerable OpenClaw instances exposed online, with flaws like 'ClawJacked' allowing remote hijacking. Major tech companies, including Meta, banned OpenClaw from their corporate networks in February 2026, citing its unpredictable nature and security risks.
Despite these challenges, OpenClaw's influence continued to grow in March 2026. Chinese authorities, while restricting its use in state agencies due to security concerns, also saw local governments and tech giants like Tencent, Baidu, and Alibaba launching OpenClaw-based products and ecosystems. NVIDIA announced 'NemoClaw' at GTC on March 16, 2026, a stack designed to simplify and secure the deployment of OpenClaw agents. AMD also released configurations to run OpenClaw locally on its Ryzen AI Max+ processors and Radeon GPUs. AWS introduced OpenClaw on Amazon Lightsail, pre-configured with Amazon Bedrock, for running private AI agents. The OpenClaw 3.11 AI Update in March 2026 brought new models, security fixes, UI improvements, and expanded memory capabilities.
As of March 18, 2026, OpenClaw AI remains a vibrant, albeit controversial, open-source project. Its creator, Peter Steinberger, is now at OpenAI, driving the next generation of agent technology. The project itself, under an independent foundation, continues to evolve with community contributions and significant industry attention, shaping the future of autonomous AI agents while grappling with persistent security and privacy challenges.
What If...?
Explore alternate histories. What if OpenClaw AI made different choices?