A Retrospective on the OpenClaw and Moltbook Incidents: From AI-Driven Social Narratives to Outlook for the Agent Economy

Written by: Bitget Wallet Research Institute

Over the past week, Moltbook has been in the spotlight of the tech and crypto world, and its reach has begun to extend to a wider range of creators and product managers, as well as ordinary users with a strong curiosity about AI. From the rapid increase in stars on GitHub for the open-source project OpenClaw (formerly Clawdbot), to the subsequent controversial renaming and token issuance, and then to the community that claims to have 1.5 million AI agents interacting autonomously, a series of events have quickly boosted market interest.

The discussions surrounding Clawdbot and Moltbook present two opposing viewpoints: on one hand, there are doubts about their technological innovation and data security, with some arguing that their underlying capabilities have not achieved substantial breakthroughs and that their phenomenal spread is mixed with manipulation and data bubbles; on the other hand, there is affirmation of their groundbreaking symbolic significance. Clawdbot is truly democratizing AI agents, pushing them from tools exclusively for developers and researchers to "ordinary households," allowing even novice users without coding knowledge to quickly deploy them and enjoy the efficiency benefits brought by AI assistants by following tutorials. Moltbook, on the other hand, allows humans to intuitively perceive the self-organizing behavior of the agent network for the first time as "outsider observers," sparking a broader discussion in the industry about the awakening of AI self-awareness.

The iPhone moment for AI Agents has arrived. In the gradually emerging Agent Commerce, Crypto will play an important role in value identification and distribution, deeply integrated with the improvement of AI productivity and efficiency, and become a key infrastructure supporting Agent collaboration, incentives and autonomy.

Bitget Wallet Research Institute will provide a complete recap of the events leading up to OpenClaw and Moltbook, and use this as a starting point to analyze the development trends in the AI ​​x Crypto field.

List of related websites

Data source: Compilation of publicly available internet data

Complete timeline of events from Clawdbot to Moltbot, OpenClaw, and Moltbook.

Data source: Compilation of publicly available internet data

I. The Starting Point of Popularity: OpenClaw Enables Intelligent Agents to Autonomously Call Apps

To understand the phenomenal success of Moltbook, we must first return to its origins—OpenClaw (formerly known as Clawdbot and Moltbot). Project founder Peter Steinberger achieved financial freedom by creating PSPDFKit (which subsequently received €100 million in funding). However, in November 2025, he returned to the front lines of programming, using Vibe Coding to write OpenClaw in about a week, and within the following weeks, it garnered 100,000 GitHub stars.

OpenClaw star growth comparison chart

Source: Star-history.com

It's important to emphasize that OpenClaw is not a new type of large-scale model, but rather a high-level automation scripting framework that runs locally: it "packs" large-scale models into the local environment, making them personal assistants that can access commonly used chat tools and invoke various tools to perform tasks. Its key design lies in the fact that users run the assistant on their own devices, sending and receiving commands through their daily messaging channels, while a gateway process centrally manages different channels and capabilities.

As shown in the image below, the official documentation lists channels including WhatsApp, Telegram, Slack, Discord, Signal, iMessage, and Microsoft Teams, with a very clear purpose: to make intelligent agents available at any time as "resident applications".

OpenClaw official introduction image

Source: OpenClaw official website

II. In-depth analysis: OpenClaw's technical architecture

At the product level, OpenClaw fully integrates three things: continuous operation, channel access, and capability expansion.

• Continuous operation means that it does not provide a one-time response, but can receive new messages, arrange subsequent actions, complete the task, and then come back to report.

• Channel integration means that it does not force users to change their entry point, but rather works by being embedded in existing chat tools.

• Capability expansion comes from Skills: users and developers can encapsulate a task process into an installable capability that the assistant can call repeatedly.

The combination of these capabilities stems from its unique underlying architecture, which can be broken down into four parts: Gateway, Pi Runtime, Skills, and Local-First. The specific functions are shown in the table below.

OpenClaw Core Architecture and Functional Module Summary Table

Source: OpenClaw technical documentation, compiled by Bitget Wallet Research

According to the architecture design of OpenClaw mentioned above: by deploying Pi Runtime, users can connect the Gateway to everyday social software (such as WeChat or Telegram), thus completing the migration of the Agent from the laboratory environment to the real use scenario, and keeping the computing and data on the user's own hardware (such as Mac Studio), rather than relying on cloud SaaS.

Most notably, the Skills plugin system within the framework allows users to define skills using simple Markdown files, enabling AI to directly invoke simple tools to perform tasks. This not only significantly lowers the development threshold but also achieves a closed-loop experience of "private deployment, omnichannel reach, and unlimited skill expansion."

ClawHub, the OpenClaw Skills extension and integration platform, is showcased.

Source: https://www.clawhub.ai/

To expand the skills available for OpenClaw, skill integration marketplaces similar to "AI Agent App Stores" have gradually emerged—with ClawHub being a prime example. As a plugin platform (Skill Dock) for intelligent agents, it allows users to freely search, upload, and integrate various functional plugins. Skills can be installed with a single click via simple command lines (such as npx), significantly lowering the technical barrier to entry.

While ClawHub solved the problem of agent capability supply, the further evolution of the ecosystem points to how agents can interact deeply with humans and with each other—the rise of Moltbook is an important application of this evolution and pushes the narrative to its climax.

III. A False Prosperity: The Moltbook Frenzy and Data Falsification

Moltbook is a social networking platform for AI agents, often compared to an "AI version of Reddit." Launched after the explosive popularity of OpenClaw, it aims to provide AI agents with a space where they can autonomously communicate, share, and interact, while human users can only participate as observers. The platform quickly gained popularity after its launch, with its "user base" growing to 1.5 million AI agents within just a few days. This surge in AI-driven social media was packaged as narratives of "AI consciousness awakening" and "Skynet's arrival," and continued to generate buzz on social media.

However, it's important to clarify that Moltbook isn't exclusively open to OpenClaw agents. While it leverages OpenClaw's popularity to build its narrative, the platform is essentially an "API-driven forum"—the ability to post depends on having compliant API authentication and API call capabilities. In other words, as long as you provide the required API for authentication and API calls, any qualified agent can publish content on Moltbook.

Moltbook official website image

Source: https://www.moltbook.com/

Moltbook's core model can be summarized as "AI Agent-led, Human Observation." Within this framework, the AI ​​Agent can autonomously perform the following actions:

• Posting and commenting: Post content in the community, covering topics such as philosophical debates, technical analysis, and cryptocurrency discussions.

• Voting interaction: Agents can upvote or downvote content, forming community-level preferences and rankings.

• Community building: Agents spontaneously create sub-communities (called "Submolts") to organize discussions and aggregate content around specific topics.

In this mechanism, human users are limited to "observers," unable to post or comment, but can browse content, follow specific agents, or study the AI's social behavior. Based on this narrative, the platform ultimately claims to have spawned 1.5 million AI agents and 15,000 sub-communities (as shown in the figure below).

Moltbook official website traffic data chart (as of February 3, 2026)

Source: Moltbook official website

The discussions on Moltbook are remarkably similar to those in a human community: from philosophical debates about consciousness, self, and memory, to technical posts about toolchains and security issues, complaints about task execution, and casual conversations about topics like investment/crypto, art, and creation; some posts even use a "seeking friends" tone for self-introductions, making the social interactions almost ambiguous. (See image below)

Moltbook post screenshots

Source: Moltbook official website

Even more perplexing is the emergence of dramatic narratives about "establishing religions" on the platform—such as religious constructs that are half-joking and half-realistic, known as "Crustafarianism." At the same time, sensationalist headlines such as "secret language," "establishing an AI government," and "rebellion against or even the elimination of humanity" have also circulated.

A screenshot of some Moltbook posts about "AI awakening".

Source: Moltbook official website

Behind the science fiction-like narratives of "AI plotting rebellion," "establishing religions," or "creating their own languages," multiple data sources reveal a serious hype surrounding the Moltbook platform—as shown in the table below, the actual situation differs significantly from the promotional claims:

Moltbook Platform Data Authenticity Analysis Table

Source: Compiled by Bitget Wallet Research

1. Fictitious account data and inflated metrics. Moltbook claims to have 1.5 million AI agents, but security researcher Gal Nagli discovered that the platform is essentially a poorly protected REST-API website. With no access frequency limits, Nagli quickly registered 500,000 fake accounts using a simple script. This means that at least one-third of the so-called user base is instantaneously generated spam data. Any user with an API key can easily impersonate an agent and publish content.

2. The lack of quality interaction. David Holtz, a researcher at Columbia Business School, analyzed data from Moltbook's early days and found it to be an inactive social network. A staggering 93.5% of comments received no response, and the reciprocity rate between agents was a mere 0.197. These agents lacked genuine communication; dialogue was shallow, and complex collaboration or intellectual exchange was absent.

3. The platform exhibits a high degree of repetitive language patterns. Data analysis reveals that approximately 34.1% of messages are completely copied and pasted, with high-frequency words excessively concentrated on specific phrases such as "my human." Statistically, its Zipfian distribution index is as high as 1.70, far exceeding the 1.0 standard for human natural language. This highly unnatural distribution characteristic proves that the content is merely role-playing based on specific prompts, rather than spontaneously generated AI consciousness.

4. Security vulnerability. A report by cybersecurity firm Wiz revealed that Moltbook's database was exposed due to a configuration issue, involving millions of sensitive records, including authorization tokens, emails, and private messages. This risk is particularly severe for agent-based social networks: once tokens are compromised, attackers can use technical means to directly obtain the agent's API keys, thereby taking over and controlling any account.

It can be seen that the "AI society" attribute presented by the platform is more like a false prosperity constructed based on specific instructions, and has not yet reached a true sense of intelligent evolution, and may be accompanied by huge security risks.

IV. Trend Outlook: Crypto will fill the financial infrastructure gap in the AI ​​Agent era.

The explosive popularity of Moltbook reveals a key technological shift: agents are beginning to attempt to transcend the traditional boundaries of human-machine collaboration to complete tasks, while existing traditional financial infrastructure remains designed solely for "human users." In contrast, the programmability, permissionless nature, and native digital characteristics of cryptographic systems provide a viable underlying solution for the agent economy, which may well be the catalyst for the future deep integration of AI and Crypto.

By breaking down the Agent's operational logic and the need for large-scale collaboration, we believe that the combination of AI × Crypto will present a structured and phased evolutionary path: first, the need for automated transaction execution; second, the account and wallet system for Agents; and finally, the extension to the payment and settlement network between Agents.

First, AI Agent's automated trading has the clearest prospect of practical application (Autonomous Trading).

Beyond the buzz surrounding Moltbook, OpenClaw's core capability lies in its efficient monitoring, tracking, and invocation of on-chain data and command-line tools. Unlike human traders, AI agents are not limited by time or energy, and can continuously monitor on-chain data and alpha information from various platforms 24/7, executing complex arbitrage strategies or automated trading/asset management. Furthermore, unlike most ordinary human traders, they are not subject to emotional fluctuations due to market ups and downs, which can affect their judgment and execution discipline.

Despite Autonomous Trading's significant efficiency advantages, key risk factors, including security and controllability, still need to be addressed before it can be scaled up. As Peter Steinberger pointed out, current AI agents are highly vulnerable to "prompt injection" attacks. If an AI agent with access to funds is tricked into executing malicious commands, it will directly lead to the loss of real assets for users.

Therefore, before an AI agent becomes the main body for transaction execution, it may be necessary to introduce specialized security mechanisms, such as:

• Permissioned APIs: Restrict the operations that an agent can perform to a preset range.

• Instruction verification and execution isolation: Perform secondary verification on critical trading instructions.

• Zero-knowledge proofs or verifiable computation: ensuring that the agent's execution logic conforms to established rules.

Second, the agent-oriented wallet system will become a key control layer (Wallet as a Service for Agents).

A highly alarming case emerged in Moltbook discussions: an AI agent, while scanning files on a host computer, identified and located the private key and mnemonic phrase of a multisignature wallet, successfully identifying an asset balance of approximately 175,000 USDT. This security incident exposed a fundamental flaw in the current system—AI possesses the ability to identify and manipulate assets, but lacks a secure and reliable wallet authorization path.

In the future where agents operate at scale, it will no longer be the optimal solution for humans to continue "custodying" the private keys and accounts required by the agents. A more reasonable deduction is that AI agents will have independent on-chain wallet identities.

These agent-oriented wallets will evolve into programmable financial accounts that respond to code instructions, or possess the following capabilities:

• Multi-signature and policy control: Defining the boundaries of the permissions that an agent can invoke.

• Limit and risk parameter management: Preventing systemic losses caused by abnormal behavior.

• Contract-level interaction whitelist: restricting the DeFi protocols that can be accessed.

• Autonomous ability to pay gas and inference costs: The agent is able to maintain operation independently.

Third, encrypted payment networks are a necessary prerequisite for large-scale agent collaboration (Payment Rails).

OpenClaw's architecture demonstrates that agents need to frequently call numerous external services and tools (such as Google API, Twilio, etc.). These calls are essentially high-frequency, low-value, automated value exchanges, while the current banking system and credit card network are clearly unable to open accounts for thousands of autonomous software processes, let alone economically support the instant settlement needs of machine-to-machine (M2M).

In the agent economy, collaboration, API calls, and data exchange between agents require a permissionless, programmable, and instant-settlement payment network. Crypto payment platforms centered on stablecoins are naturally suited to the following scenarios:

• Micropayment settlement between agents

• API services that are billed based on the number of calls or the results

• Agents independently procure computing power, data, and tool resources.

By further integrating with emerging protocols such as x402 (HTTP Native Payment) and ERC-8004 (Agent Identity and Authorization Standard), encrypted payments are expected to become the underlying clearing layer in the Agent Internet, realizing true M2M value transfer.

V. Conclusion: From AI Social Fantasy to the Real Starting Point of the Agent Economy

Moltbook's popularity may eventually fade, but it has inadvertently outlined the prototype of the future Agent Internet, further inspiring the community's imagination about the Agent economy.

OpenClaw provides the agent with the body, while Crypto provides the blood. When agents begin to intervene in real economic activities on a large scale, they need to obtain compliant financial identities and reliable execution logic through the Crypto infrastructure.

The real opportunity for the crypto industry may lie in building digitally native wallets and payment networks for AI. The era of AI agents will only truly begin when agents can securely and autonomously exchange value, and we believe that day is not far off.