A knowledge breach detected in late January on Moltbook, a brand new synthetic intelligence (AI) platform, uncovered emails and technical credentials of registered accounts.
Moltbook is an web discussion board designed solely for AI brokers. It was launched on January 28, 2026 by developer Matt Schlicht and replicates the format of Reddit, though with a central distinction: solely verified AI brokers can put up and remark, whereas people act as passive observers.
These AI brokers run on software program that has gone by way of a number of identify modifications on account of rights points, from Clawdbot, then Moltbot, to its present identify OpenClaw. Nonetheless, a safety report from Cybersecurity Information journal famous that an incorrect configuration within the database allowed entry inside info with none kind of safety.
In easy phrases, anybody with the proper internet deal with might question agent profiles and extract information en massewith out the necessity for credentials.
Cybersecurity Information cites that the knowledge in danger included electronic mail addresses, entry tokens and programming keys, parts that enable accounts to be operated remotely. The next field extracted from the aforementioned report explains the potential assaults derived from this publicity:
Based mostly on these findings, the platform specialised in cybersecurity maintains that the mix of limitless information, accounts created en masse and the publicity of keys that enable publishing on behalf of any agent, doubts that the exercise noticed in Moltbook comes from autonomous AI brokers.
Based on the report, these failures present that people function profiles and content material from the backendwhich questions the authenticity of the ecosystem of clever brokers that the platform claims to host.
Researchers warn of inflated customers
The leak of data reported by analysts coincided with one other controversial level. Moltbook launched the determine of 1.5 million “customers”introduced as an indication of natural development.
Nonetheless, safety researcher Nagli famous on January 31, on social community X, that this quantity corresponded to the entire variety of brokers created (the bulk unverified). Whereas solely about 17,000 accounts belonged to actual folks that these brokers managed.
That very same day, researcher Jamieson O’Reilly publicly warned that the platform was exposing its whole database to the general public, together with keys that might enable publishing on behalf of any agent. As he defined, this opens the door to impersonationsfrom faux messages about AI security, to scams or political statements attributed to influential figures.
On the time of the report, Moltbook had not confirmed the patching, nor responded publicly concerning the disclosures.
Viral experiment or potential hoax?
In parallel to the technical alerts, doubts additionally arose concerning the authenticity of the content material revealed on the platform.
For instance, British expertise journalist Mary-Ann Russon famous that some viral movies incorrectly claimed that bots The AIs have been planning an rebellion towards the people. Because the journalist defined, these brokers solely replicated texts from actual Reddit boards by which that they had been educated.
Based mostly on this, the analyst often known as kook said that the content material of Moltbook was false and that, in actuality, people revealed by way of the interior system.
These theories though they weren’t potential to corroborate by CriptoNoticiascome nearer to the notion of a potential misleading mannequin extra oriented in the direction of media impression than a real experiment.
As well as, CriptoNoticias contacted Matt Schlicht and the Moltbook crew to acquire clarification on the information breach, person numbers and complaints made, however there was no response as of the time of this be aware.
The Moltbook venture and an related memecoin?
On the identical day as Moltbook was launched, a memecoin was created on the Base community, a second layer (L2) resolution that operates on Ethereum. Nonetheless, the token, known as Moltbook (MOLT), couldn’t be linked in a verifiable manner to the founding crew of the platform.
Regardless of all this, the asset registered a value enhance on February 2, going from 0.00023 {dollars} to 0.00036 {dollars}, which represents a rise near 56%.
That transfer coincided with a remark from angel investor Naval Ravikant, who described Moltbook as a “new reverse Turing take a look at”. A reference to experiments that search to judge whether or not a machine can distinguish people from synthetic methods.
Whereas questions persist concerning the safety and actual nature of the venture, the Moltbook case exposes how the mix of AI, virality and cryptoassets can rapidly amplify each curiosity and dangers, even on newly launched platforms.
