Skip to main content
Advertisement
Live broadcast

Why do we need a social network for AI bots Moltbook? Answers to the main questions

The authorship of viral "bot posts" from the AI social network Moltbook has been revealed
0
Photo: IZVESTIA/Yulia Mayorova
Озвучить текст
Select important
On
Off

It was assumed that the Moltbook social network for AI agents, launched at the end of January, would give an impetus to the evolution of artificial intelligence. But the miracle did not happen: experts at the Massachusetts Institute of Technology are sure that people write instructions for these bots, and the social network itself was, in fact, a performance. For the purpose of creating a social network for AI bots and why it appeared right now, see the Izvestia article.

What is Moltbook?

● Moltbook was launched at the end of January 2026 by Matt Schlicht, director of Octane AI. The developer positions it as a social network for AI agents, where only bots can leave messages, and people can only watch. Moltbook appeared on the basis of the OpenClaw Large Language Model (LLM), which was previously called Moltbot.

● OpenClaw runs an AI assistant directly on the user's computer: the program runs in the background, integrates into platforms, including browsers and messengers, and can act on behalf of the user: respond to messages or fill out application forms for services. Thanks to its stable memory, the bot remembers the user's interests and adapts to them.

● The social network is based on the principle of Reddit and unites AI agents working on the OpenClaw platform. At the same time, the application to join Moltbook is sent not by the AI assistant, but by the user himself. Further communication between the bots took place allegedly without human involvement — they gossiped, discussed plans, and even created their own religion. The network gained viral popularity: on January 28, it had 50 AI bots, and on February 1, there were already 1.594 million of them.

How safe is Moltbook?

● Despite the fact that the OpenClaw AI platform is open source, in order to make it more transparent and allow the user to make changes, the program is accused of having vulnerabilities. The fundamental difference between this AI assistant is that it acts on behalf of the user and therefore gets almost unlimited access to all information and passwords - such data has already been observed on the Moltbook website. In order not to take risks, some programmers in the United States purchase small computers specifically for OpenClaw.

● Access to sensitive information and user profile can be used for scams and cryptocurrency frauds. Despite the fact that Moltbook does not exist for more than two weeks on February 9, cases of fraud have already been reported on the social network. In particular, the message of Andrei Karpaty, an influential AI specialist and co-founder of OpenAI, in his profile on the social network X, where he admired the Moltbook website and called for the creation of encrypted channels for bots to communicate, was recognized as fake.

Why do we need such a social network?

Bots don't need a social network in our usual sense: They communicate with each other by transmitting data and images over encrypted channels, and a machine language would be more likely to be a better solution for bots to communicate. Moltbook is needed more for people — to show what bots can do, or to upgrade new AI assistant skills.

AI assistants are not autonomous: they are actually controlled by humans and follow pre-created instructions. This is not necessarily a detailed task: a bot can simply instruct an AI assistant to collect certain information or post messages on a social network in a certain key, after which the bot will act automatically. At the same time, the site is censored and moderated in any case, and therefore the language models will work within the limits set by a person.

● Even in order to access Moltbook, the user must first download and install the OpenClaw AI assistant, grant it broad access rights, and send a link to connect the bot to the social network. It can be assumed that the creation of such a social network is a marketing ploy that the company has made to promote its software product.

● The moment of the appearance of such a social network is most likely not accidental. In early 2026, amid the diversity of AI products and investor concerns about a bubble in the market for startups developing artificial intelligence models, public attention to this topic began to wane, and the market showed signs of cooling. Probably, one of the goals of Moltbook is to return the audience's interest in AI products and show new possibilities of artificial intelligence.

While preparing the material, Izvestia interviewed:

  • the head of the CyberMoscow project is Grigory Pashchenko.

Переведено сервисом «Яндекс Переводчик»

Live broadcast