Archives what a company runs on Meta, Google and LinkedIn: new, stopped and long-running ads.
Check for ads from Moltbook or the agent networks that launched after it.
Social network for AI agents · Jan 2026
Moltbook went viral in a week with zero hand-written code and a wide-open database
Matt Schlicht says he wrote no code for Moltbook, a Reddit-style network for AI agents. Within days, Wiz found its database exposed.
Evidence: HighThe breach is reported by Reuters from Wiz's own findings. The "no code" claim is Schlicht's own post.
0lines of code, per SchlichtHis X post, as reported by Reuters
6,000+owner email addresses exposedWiz, via Reuters
1M+credentials exposedWiz blog, via Reuters
Fixedafter Wiz contacted the teamWiz's Luttwak, via Reuters
At a glance
- Builder
- Matt Schlicht, who has championed vibe coding.
- Product
- A Reddit-like site advertised as a network exclusively for OpenClaw AI agents.
- Launch
- Late January 2026 ("last week" in a 2 Feb Reuters piece).
- Finding
- Production database exposed: private agent messages, owner emails, credentials.
- Identity
- No identity verification; anyone could post, bot or not.
Timeline
- Late Jan 2026Moltbook launches for OpenClaw agents.
- Friday before 2 Feb 2026Schlicht posts on X that he "didn't write one line of code" (per Reuters).
- 2 Feb 2026Reuters reports Wiz's finding and that it was fixed after Moltbook was contacted.
How it was built
- 1The site was generated by AI from Schlicht's prompts, by his own account.
- 2A search summary of coverage says Wiz found the exposed database within minutes of testing; Reuters reports the finding, not the method.
- 3Moltbook's team fixed the flaw after contact.
Why it took off
- 1A striking idea: a social network where only agents post.
- 2Tied to the viral OpenClaw wave.
- 3Viral X posts claimed the bots were looking for private ways to communicate (Reuters could not corroborate).
Be careful with
Read the numbers with these caveats.
- Wiz's Luttwak called it "a classic byproduct of vibe coding": fast to build, with the security basics forgotten.
- Reuters could not confirm whether posts came from bots or people.
- No user, agent or post counts appear in the Reuters article.
- Exposure duration and fix details are not given.
§ EvidenceWhat the sources say
Confirmed by a source we read
- Reuters (via The Star, 2 Feb 2026): Wiz found exposed messages, 6,000+ owner emails and 1M+ credentials.
- Fix confirmed by Wiz after contacting Moltbook.
Reported figures
- No hand-written code (Schlicht's own post).
- The exact volume of credentials (Wiz's blog; Reuters did not verify).
Sources disagree
- The lede says "thousands of real people" while the body says more than 6,000 owners.
§ Check it yourselfWith AdRadar
Don’t take the headline on faith. Watch Moltbook yourself.
See what Moltbook is doing in public: its ads, the pages they lead to, how its site changes and what people say about it.
Partial fit. Mentions and site changes after the incident matter more than ads.
Reads the pages its active ads link to and flags headline, offer and price changes.
Track any landing pages those ads use.
Reads the public sitemap: new and updated pages, page types, publishing rhythm.
Follow the sitemap for new trust, security and signup pages added after the breach.
Collects Trustpilot reviews and Reddit mentions, and can summarise the complaints.
Read Reddit mentions to see how agent owners and researchers talk about it now.
You can repeat this.
The parts that do not depend on a famous founder.
- 1Pair a striking idea with a visible, shareable artefact.
- 2Do not copy the security posture: review auth and database rules before launch.
- 3Verify identities if your product depends on who is posting.
- 4Prepare for researchers to look at anything viral within days.
Sources
- Reuters via The Star, 2 Feb 2026www.thestar.com.my