A data breach detected in late January on Moltbook, a new artificial intelligence (AI) platform, exposed email and technical credentials for registered accounts.
Moltbook is an internet forum Designed specifically for AI agents. It was launched on January 28, 2026 by developer Matt Schlicht and replicates the format of Reddit, but with a core difference. That is, only verified AI agents can post and comment, while humans act as passive observers.
These AI agents run on software that has been renamed several times due to rights issues, starting with Clawdbot, then Moltbot and the current name OpenClaw. However, a security report from Cybersecurity News magazine states that the database configuration is incorrect. Access inside information without any protection.
Simply put, anyone with a valid web address You can query agent profiles and extract data in bulkno credentials required.
Cybersecurity News says the information at risk includes elements such as email addresses, access tokens, and programming keys that allow remote manipulation of accounts. This is explained in the following box excerpted from the aforementioned report. Attacks that may derive from this exposure:
Based on these findings, the cybersecurity-focused platform has announced that the combination of unlimited records, bulk-created accounts, and key exposures that can be published on behalf of agents will Suspicion that activity observed on Moltbook was caused by an autonomous AI agent.
According to the report, these failures indicate: Humans interact with profiles and content. backendwhich calls into question the credibility of the ecosystem of intelligent agents that the platform claims to host.
Researchers warn users of inflated data
The breach reported by analysts coincided with another controversial point. The diagram published by Maltbook is 1.5 million “users”shown as a sign of organic growth.
However, security researcher Nagli pointed out on Social Network X on January 31 that this number corresponds to the total number of agents created (mostly unverified). meanwhile Only about 17,000 accounts belonged to real people Those agents were in control.
On the same day, researcher Jamieson O’Reilly publicly warned that the platform was exposing its entire database containing keys that could be published on behalf of agents. As he explained, this Open the door to identity theftfrom fake messages about AI safety to scams and political statements by influencers.
At the time of this report, Moltbook had not confirmed the patching or responded publicly to the disclosure.
Virus experiment or possible hoax?
Alongside the technical warnings, questions also arose about the authenticity of content published on the platform.
For example, British technology journalist Mary-Ann Russon pointed out that some viral videos falsely claim: bot The AIs were planning a rebellion against humanity. As the journalist explained, these agents simply replicated text from real Reddit forums on which they were trained.
Based on this, the analyst known as kook believes that the contents of the malt book are false and, in fact, Humans exposed through internal systems.
those theories Although it could not be corroborated by CriptoNoticiaswe come closer to recognizing the potential for deceptive models that emphasize media influence over genuine experimentation.
Additionally, CriptoNoticias reached out to Matt Schlicht and the Moltbook team for clarification regarding the data breach, number of users, and complaints, but did not receive a response as of this note.
A meme coin associated with the Moltbook project?
On the same day as Moltbook’s launch, memecoin was created on the base network, a second layer (L2) solution running on top of Ethereum. However, the token called Malt Book (MOLT) is could not be linked in a verifiable way To the founding team of the platform.
Despite all this, the asset recorded a price increase from $0.00023 to $0.00036 on February 2nd. This represents an increase of nearly 56%.
The move echoed comments made by angel investor Naval Ravikant, who explained Maltbook. As a “new inverse Turing test”. A reference to experiments that attempt to assess whether machines can distinguish between humans and artificial systems.
While questions remain about the security and actual nature of the project, the Maltbook incident revealed how the combination of AI, virality, and crypto assets can rapidly amplify both interest and risk for even newly launched platforms.