The OpenClaw Story: A Wild Turn in the AI Agent Revolution

Imagine building a simple weekend project that explodes across the internet, attracts cryptocurrency scammers, triggers trademark threats, and ultimately culminates in an offer from OpenAI. This is exactly how the story of Peter Steinberger and his OpenClaw — formerly known as Claudebot and Moltbot — unfolded. At AI w Biznesie, we have tracked this saga from the beginning, and it is undoubtedly one of the wildest journeys in the world of AI. Based on recent developments and insider sources, here is the full story of how a side project reshaped the AI landscape.

• The Birth of a Phenomenon: From Side Project to Viral Sensation
• Trademark Drama and the Security Crisis
• The Strategic Pivot: Why OpenAI Recruited Peter Steinberger
• The Future of AI Agents: What This Means for Business

In the tech world, it is rare to see a story that combines brilliant code, corporate blunders, hacker attacks, and high-stakes recruitment within just a few weeks. The story of OpenClaw is a fascinating case study of the breakneck speed at which the AI sector evolves. It is a tale of how one developer forced the industry’s biggest players — from Anthropic to OpenAI — to redefine their entire strategies regarding AI agents.

The Birth of a Phenomenon: From Side Project to Viral Sensation

It all began in November 2025 as a modest side project. Peter Steinberger, an Austrian developer living in Silicon Valley and the founder of PSPDFKit (a PDF toolkit used by Apple, Dropbox, and SAP), created a simple „glue” script. It connected the API of Anthropic’s Claude model with WhatsApp. His goal was convenience: Peter wanted to send commands to his AI via the messaging app he used daily. However, what set his project apart from thousands of other chatbots was that it wasn’t just a conversational interface. It was an autonomous agent.

An Agent That Actually Works

Unlike standard interactions with ChatGPT or Claude, where the user receives only a text response, Steinberger’s project could take action. Thanks to integrations with various services, a user could write: „Book me a table at a restaurant” or „Check my inbox and summarize the most important emails,” and the agent would actually log into the relevant systems to perform the task. This transition from conversational AI to operational AI became the foundation of the project’s success, originally named Claudebot.

Explosion on GitHub

By January 2026, the project went viral on an unprecedented scale. In just one week, the GitHub repository amassed over 200,000 stars, becoming the fastest-growing open-source project in the platform’s history. The site recorded over 2 million unique visits per week. An entire ecosystem emerged around it, including Moltbook — a social network for AI agents where bots could exchange information and learn from one another. Even Andre Karpathy, co-founder of OpenAI, described the phenomenon as the most „sci-fi” thing happening in reality lately.

At AI w Biznesie, we have long emphasized that the true value of artificial intelligence lies in process automation, not just content generation. The success of OpenClaw confirmed this thesis on a massive scale. Users don’t want another chatbot; they want a digital assistant that relieves them of repetitive tasks. Steinberger provided the code that made this possible, which immediately caught the attention of both enthusiasts and corporate lawyers.

Trademark Drama and the Security Crisis

The sudden popularity of the project brought unexpected challenges. The first was a conflict with Anthropic, the creator of the Claude models. In late January 2026, Anthropic’s legal team sent Peter Steinberger a cease-and-desist letter regarding trademark infringement. The name „Claudebot” was deemed too similar to their flagship product. While legally defensible, this move triggered a chain of events that nearly collapsed the project and ultimately shifted the balance of power in the AI market.

Scammer Attacks and „Special Operations”

Peter Steinberger agreed to a rebrand, choosing the name Moltbot (a play on the molting of a lobster, Claude’s unofficial theme). However, the moment he released the old username on GitHub and Twitter to claim the new one, cryptocurrency scammers pounced. Within seconds, they hijacked the old accounts, promoted fake tokens on the Solana network, and served malware to unsuspecting users. It was a critical moment — Steinberger considered deleting the project entirely. Instead, he executed a secret operation: a second rebrand to OpenClaw, planned with military precision to avoid further hijacking. He described the stress of those 10 hours as his personal „Manhattan Project.”

The Cybersecurity Nightmare

Alongside branding issues came real technical threats. Cybersecurity firms, including Gartner and CrowdStrike, began flagging „unacceptable cyber risks.” Researchers discovered over 30,000 instances of OpenClaw publicly accessible on the internet without any authentication. Each of these instances had access to private user data: emails, calendars, Slack tokens, and API keys. It turned out that 93% of verified installations had critical vulnerabilities. It was a paradox: the world had received its most useful AI agent, which simultaneously acted as a massive hole in data security systems.

This stage of the OpenClaw story serves as a vital lesson for any company implementing AI automation. At AI w Biznesie, we always stress that speed of implementation cannot come at the expense of security. The OpenClaw example shows that even a brilliant tool can become a liability if it is not properly secured within a corporate architecture. These security hurdles became a primary reason the project needed a powerful technological partner.

The Strategic Pivot: Why OpenAI Recruited Peter Steinberger?

Despite the security flaws, the potential of OpenClaw was too significant for tech giants to ignore. Steinberger, having made a fortune from the sale of his previous company (PSPDFKit sold for approximately 100 million euros), was funding the project’s infrastructure out of his own pocket, costing him $10,000 to $20,000 a month. He knew the scale was becoming unmanageable. Soon, the offers started pouring in. Mark Zuckerberg of Meta expressed interest, and Satya Nadella of Microsoft reportedly called the developer personally. Ultimately, Sam Altman and OpenAI won the battle.

Why OpenAI?

The choice of OpenAI was not accidental. From Steinberger’s perspective, the company offered the best conditions to maintain the open-source nature of OpenClaw while providing massive computational resources and expertise. For OpenAI, hiring Steinberger was a masterstroke in the battle for the enterprise market. Data from mid-2025 showed a worrying trend for Altman’s company: their share of the enterprise API market had dropped from 50% to 25%, while Anthropic grew to 32-40% thanks to the success of tools like Claude Code.

Capturing the Competitor’s Ecosystem

The most ironic part of this story is that OpenClaw was generating massive revenue for… Anthropic. Most users running OpenClaw needed paid Claude API subscriptions for the agent to function. By recruiting the creator of OpenClaw, OpenAI didn’t just gain a brilliant engineer; they effectively „cut off the oxygen” to their biggest competitor by taking control of the world’s most popular agentic ecosystem. Sam Altman announced that OpenClaw would move into a foundation supported by OpenAI, ensuring its continued development as an open-source project.

At AI w Biznesie, we analyze this as a pivotal moment in the „Agent Wars.” The battle is no longer about whose large language model (LLM) scores higher on benchmarks. The fight has moved to the agent layer — the software that sits between the model and the user and can actually perform work. By integrating Steinberger’s vision, OpenAI aims to create a standard for AI agents that is so simple to use that anyone, not just developers, can master it.

The Future of AI Agents: What This Means for Business

Peter Steinberger joining OpenAI and the transformation of OpenClaw into a supported project signals that the era of „talking chatbots” is coming to an end, and the era of „acting assistants” is beginning. This paradigm shift has immense implications for how we work and manage businesses in the coming years.

The Agentization of Daily Life

We can expect that the features made famous by OpenClaw will soon be directly integrated into ChatGPT. Steinberger’s vision — an agent that „even his mom could use” — suggests that OpenAI will prioritize radical simplicity. Instead of complex GitHub configurations, we will likely see a button in the app that allows the AI to manage our calendars, pay bills, or organize business trips. This is the democratization of advanced automation, previously reserved for those with technical expertise.

Security as a Priority

The biggest challenge for OpenAI will be repairing OpenClaw’s reputation regarding security. For businesses to fully trust AI agents, they must have absolute certainty that their data is protected. We can expect OpenAI to introduce new standards for authentication and data isolation for agents, which could become a new industry benchmark. At AI w Biznesie, we predict that security will be the primary field of innovation in 2026.

A Lesson for Brands: Unintended Consequences

This story is also a warning to companies against overly aggressive protection of intellectual property at the expense of community goodwill. By sending a legal threat to Steinberger, Anthropic protected its brand but lost its most valuable ambassador and a massive stream of API revenue, which landed straight in the hands of OpenAI. In the new AI economy, collaborating with open-source creators often yields more benefits than rigid adherence to corporate procedures.

In conclusion, the OpenClaw story is just the beginning. We are witnessing the birth of a new category of software. For entrepreneurs and business leaders, the message is clear: it is time to stop viewing AI as a tool for writing text and start seeing it as an autonomous employee. At AI w Biznesie, we help companies prepare for this revolution by implementing secure and effective automations that leverage the full potential of AI agents. The future will be „multi-agent,” and those who are first to learn how to manage them will gain a competitive advantage that will be difficult to overcome.