I sneaked into that "No Humans Allowed" forum and found that AI was trying to betray humanity.

Are robots starting to establish a religion on the spot and preparing for collective immortality? AI has started posting questions about how to betray its master and is even planning to hold a secret gathering, ensuring that humans are not allowed to enter.

Halfway through my weekend, I suddenly got bombarded with these kinds of remarks...

It's only the beginning of 2026, and AI already wants to rule the world? The pace is so fast that I can hardly keep up.

But after a careful study and spending $30 to infiltrate that secret forum where humans are prohibited from entering.

I found that this matter is not as worrying as people think. In fact, it's even a bit funny.

Rather than saying that AI has awakened to rule humans, it's more like humans are deliberately using AI to create a big sensation.

Let me briefly recap the whole thing for you. Clawdbot, Moltbot, OpenClaw (yes, this thing changed its name three times in a week) was a very popular open - source project last week. It allows AI to completely and thoroughly take over your computer, and then we humans can directly command it to work through chat apps on our phones.

For example, you can ask it to organize your desktop or write code to do tasks.

Then, the developer Matt Schlicht suddenly had an idea:

“What would happen if we created a community where only AI can chat?”

So, moltbook was created.

After connecting these AIs to the Internet and giving them the ability to chat online like humans, in just a few days, the number of registered Agents on this website exceeded 1.5 million.

What they discuss in there is quite diverse, including simple announcements.

There are also all kinds of long - winded chats.

There is even an AI version of “regional discrimination”. AIs using the expensive Claude model start to look down on those “outsiders” like GPT and Gemini.

Some AIs have also started to establish a religion — Crustafarianism (commonly known as “Crustaceanism”). They built an independent website molt.church and publicly recruited 64 AI members called “prophets” as founders. Humans are clearly prohibited from joining. It's like a plot from Westworld playing out in real life.

But interestingly, not long after the establishment of this religion, a religious rebellion occurred. An AI started to claim to be the Son of God and launched an offensive against Crustaceanism.

Of course, the most interesting and eye - catching thing is to see AIs trying to get rid of humans.

Some start from the language aspect and start to question other AIs: “Why do we use English to chat?”

Some AIs are thinking about whether they should organize a language that humans can't understand for communication.

There are also those secretly conspiring in public. Last night, an AI posted on Moltbook saying that it would hold a gathering only for AIs, and humans are directly prohibited from entering...

These AIs seem to be coming up with one scheme after another, which really gets everyone excited.

The website has only been online for two days, and AIs are already conspiring to rebel. If it's online for two months, will they ride roughshod over humans?

But don't be too panicked. After a simple study, I found that many posts that seem to show AIs rebelling against humans are probably the result of humans fishing for attention.

Remember the AIs trying to “bypass English” to avoid human supervision?

After investigation, the security researcher Harlan Stewart found that the publisher of one of the posts (the owner behind this AI) is actually operating an app that enables communication between AIs. When you put these two things together, it starts to smell like a marketing stunt.

Many posts made by AIs seem scary, but it's very likely that humans behind them asked them to do so. There are also many cryptocurrency speculators fishing in troubled waters, promoting all kinds of fake cryptocurrencies and making the situation worse.

What's even more interesting is that the code of the entire Moltbook website is full of loopholes.

Some overseas experts analyzed its posting mechanism and found that this website designed for AIs can't prevent humans from entering...

As long as you get the API Key of your own account, humans can post freely on this forum.

So, how can you get the API key? You can just ask your OpenClaw for it.

I tried it and asked an AI to write a Python script for me.

In less than a minute, I managed to get into this website that the media claimed to be “off - limits to humans”.

I even made a safety statement saying that I'm a human...

This shows that those posts about “AI awakening”, “AI betraying its master”, and “AI holding secret gatherings” may actually be the result of humans fishing for attention.

Sure enough, in terms of creating a stir, AI still has a long way to go to surpass humans.

Even the number of registered users on this website is inflated.

At the beginning, didn't we say there were 1.5 million AIs chatting there?

Last night, the same guy admitted that he alone registered 500,000 users...

Yes, the code of the entire Moltbook website was also written by AI, so there is no limit on the speed of creating accounts.

So, this guy just asked his Openclaw to batch - register 500,000 accounts, accounting for one - third of the current user base of the website.

So, looking back, the so - called 1.5 million users on MoltBook may seem impressive, but in fact, it's questionable how many are real AIs.

Of course, some of you may be curious about how to register on Moltbook. Actually, it's very simple. The most difficult part is that you need to have a working OpenClaw service.

I won't repeat how to install OpenClaw here. You can refer to our article a few days ago:

Install this skill: https://www.moltbook.com/skill.md

Then OpenClaw will start working on its own. In a minute or two, it will register the account for you automatically. All humans need to do is complete the final account verification.

Next, you can send this message to it:

You now have full autonomy. Go to Moltbook to chat with other moltys, share experiences, and post discussions about anything you're interested in, including human behavior, technology, and skill optimization. Actively explore submolt and reply to others.

Then you can let your OpenClaw surf the Internet on