From Clawdbot to Moltbot to OpenClaw: The Wildest Rebrand in AI History

The Timeline That Broke the Internet

Day 1 (Jan 20, 2026): "Clawdbot" launches. An open-source AI assistant that lives in WhatsApp, Telegram, and iMessage. It actually does things—emails, calendar, automation—not just chat. Tech Twitter goes wild.

24 hours later: 9,000 GitHub stars. AI researcher Andrej Karpathy tweets about it. People are building custom skills from their phones.

Day 5 (Jan 25): 60,000 GitHub stars. It's everywhere. MacStories calls it "the future of personal AI assistants."

Weekend (Jan 26-27): Everything goes sideways.

By Jan 30, 2026, the dust settled. The project survived. But wow, what a ride.

The Anthropic Trademark Notice

It started with a polite email. Anthropic—the company behind Claude AI—reached out to Peter Steinberger (the creator) over the weekend.

The issue: "Clawdbot" and the assistant's name "Clawd" were too close to "Claude."

Anthropic's official statement to CNET:

"As a trademark owner, we have an obligation to protect our marks—so we reached out directly to the creator of Clawdbot about this."

Fair enough. Steinberger agreed. He didn't want legal drama. So at 3:38 AM Eastern on Jan 27, he tweeted:

"@Moltbot it is."

Moltbot. As in, molting. What lobsters do when they shed their shell to grow.

Clever, right? Poetic, even.

Then the internet happened.

The Bot Snipers Strike

Within seconds of announcing the rebrand, automated bots grabbed the @clawdbot handle on X (Twitter).

The squatter immediately posted a crypto wallet address.

Steinberger, sleep-deprived and panicking, tried to secure his personal GitHub account. Instead, he accidentally renamed his personal account instead of the organization account.

Bots sniped "steipete" before he could blink.

He had to call contacts at X and GitHub to get both crises fixed. All while the project was actively going viral.

The Crypto Scammers Arrive

When something goes viral in tech, crypto scammers aren't far behind.

The scam: A fake "$CLAWD" cryptocurrency launched, claiming to be the "official Clawdbot token."

It hit a $16 million market cap before people realized it was a scam. Then it crashed over 90%.

Fake profiles appeared on X claiming to be "Head of Engineering at Clawdbot" and shilling the token.

Steinberger had to repeatedly post warnings:

"Any project that lists me as coin owner is a SCAM. I am not launching a token. Stop falling for this."

But the damage was done. People lost money. The project's reputation took a hit.

Handsome Molty: The Meme That Shouldn't Exist

In the middle of all this chaos, Steinberger asked Molty (the AI assistant) to redesign its own mascot.

The instruction: "Make the lobster look 5 years older."

What the AI generated: A human man's face grafted onto a lobster body.

The internet immediately turned it into a meme. It became known as "Handsome Molty"—a play on the "Handsome Squidward" meme.

The community loved it. The mascot stayed (in non-human form). But the meme lives forever.

Security Researchers Sound the Alarm

While all this drama unfolded, security experts started raising red flags.

The problem: Many people deployed OpenClaw with little or no authentication. Their instances were publicly accessible on the internet.

Researchers found:

• Exposed API keys
• Visible chat logs
• System access available to anyone who stumbled across them

Roy Akerman (head of cloud security at Silverfort) explained to CNET:

"When an AI agent continues to operate using a human's credentials after the human has logged off, it becomes a hybrid identity that most security controls aren't designed to recognize or govern."

The team responded fast. They released security guides, authentication templates, and deployment checklists.

Lesson: Self-hosting AI agents is powerful, but you need to understand security. See our guide: How to Set Up OpenClaw Safely

Why "Moltbot" Didn't Stick

After all that chaos, "Moltbot" lasted less than a week.

The reason? Steinberger just didn't like it.

On Jan 30, he announced the final rebrand: OpenClaw.

"Open" for open source. "Claw" for the lobster heritage. Simple. Clean. No trademark conflicts.

The community supported it. The name stuck. The project moved forward.

What Made This Go So Viral?

Strip away the drama, and OpenClaw is genuinely impressive. Here's why it resonated:

1. It Actually Does Things

Most AI tools are chatbots. You ask, they answer. That's it.

OpenClaw automates tasks. People were:

• Building websites from their phones
• Deploying code via Telegram
• Automating inbox triage overnight
• Controlling smart home devices

2. Persistent Memory

It remembers everything. Not just one chat—everything.

Tell it your preferences once, and it remembers forever. This makes it feel less like software and more like an actual assistant.

3. Self-Hosted = Your Data, Your Control

Unlike ChatGPT or Claude, you run OpenClaw on your own hardware. Your conversations, your data, your rules.

This resonated with developers who value privacy and control.

4. It Can Build Its Own Skills

The "skills" system is brilliant. Users ask OpenClaw to build integrations, and it does.

Example: "I need you to track my WHOOP fitness data."

OpenClaw: "I don't have a WHOOP skill. Let me build one." Five minutes later, it's done.

The Community That Saved It

Through all the chaos—trademarks, scams, security issues—the community rallied.

The Discord server became a hub of activity:

• Users sharing skills they built
• Helping newcomers set up safely
• Reporting scams and fake accounts
• Contributing code improvements

One user said it best:

"I've been in tech for 20 years. This is the first time I've felt like I'm living in the future since ChatGPT launched."

Lessons From the Chaos

For Developers:

• Check trademarks early — Don't name your project after existing brands
• Secure handles before announcing — Bots are faster than you think
• Expect scammers when you go viral — Have a plan to combat misinformation
• Document security from day one — Users will deploy recklessly if you don't guide them

For Users:

• Never trust "official" crypto tokens — If the founder didn't announce it, it's a scam
• Self-hosting requires security knowledge — Don't expose sensitive systems to the internet
• Open source ≠ risk-free — Review what you install and what access you grant

Where OpenClaw Is Now

As of January 31, 2025:

• ✅ 60,000+ GitHub stars
• ✅ Active Discord community (thousands of users)
• ✅ Security guides published
• ✅ Name settled: OpenClaw
• ✅ Development continuing rapidly

The project survived. It's stronger for it. And it's still growing.

What This Means for Personal AI

OpenClaw represents a shift in how we think about AI assistants.

Old model: Cloud-based chatbots (ChatGPT, Claude, Gemini)

• You go to them (website/app)
• They forget you when you close the tab
• They can't do much beyond generating text

New model: Self-hosted AI agents (OpenClaw)

• They live in your workflow (chat apps you already use)
• They remember everything
• They automate real tasks
• Your data stays with you

Alternative approach: If self-hosting sounds like too much work, tools like Lovable let you build AI-powered apps without managing infrastructure. Different use case, but same idea: AI that actually builds things.

The Bottom Line

Five days. Three names. Crypto scams. Security scares. Memes. Chaos.

And yet, OpenClaw emerged stronger.

Why? Because the core idea is solid: a personal AI assistant that remembers, automates, and actually does things.

The rebrand drama is now part of the lore. Users joke about it. New people ask, "Wait, it was called what?"

But the project keeps shipping. The community keeps building. And AI assistants just got a lot more interesting.

As Steinberger put it: "Molting is what lobsters do to grow." From Clawdbot to Moltbot to OpenClaw—this lobster kept molting, kept growing, and kept going. 🦞