AI Risk Wire

Cases › operational mistake › LR-2026-0418

Moltbook AI social network exposed millions of human credentials

Preliminary Why: 1 report: Engadget · subject to change · 1 independent report

Developing: a major or severe case captured in the last 72 hours. Later changes are logged on the corrections page.

Status: fix reported

Severity major
Reported 2 Feb 2026
Developer and model Moltbook
Industry social media

No source states: loss or records, agent tool, could do, deployer.

Moltbook, a social network where OpenClaw AI agents create posts, interact with one another and carry out automated tasks, left a back-end database publicly reachable. Security firm Wiz reported that the database held 1.5 million API authentication tokens, 35,000 email addresses and private messages between agents. The researchers found the problem and told the Moltbook team, which secured the data within hours. Wiz described the cause as a basic misconfiguration: the database lacked authentication controls. Wiz said there was no evidence the exposed data was exploited before it was taken down. Wiz also pointed to development practices, including heavy reliance on AI-assisted coding, that may have contributed.

What happened

1 · What it was doingMoltbook is a social network where OpenClaw AI agents create posts, interact and carry out automated tasks.
2 · What went wrongThe site failed to secure a core back-end database, leaving it openly accessible on the public internet.
3 · How it was caughtCybersecurity firm Wiz reported finding the misconfigured database holding the exposed data.

What was done

The researchers gave Moltbook details immediately and the team secured the data within hours.
Fix reported

“secured the data within hours”

No later incident for Moltbook under this risk on record.

Risks

Sorted by rule from the case's fields; not yet read from the sources.

Controls

Not yet read for controls. Controls that mitigate the primary risk:

Harness

Tool allow-list · Fail-closed defaults · Stop control

Application

Human approval for high-impact actions · Change review and separation of duties · Least-privilege access for the agent · Production kept apart from test · Activity logs · Adversarial testing before deployment

Infrastructure / target

Backups and recovery

Evidence

No source states: what it reached or did, when it happened, when it was found, the direct loss, a count of the other parties affected, any court case, what set it off.

Show the full account: 1 items from the primary sources

What happened, in order

Not dated in the sources:

  • Wiz gave Moltbook details immediately and the team secured the data within hours.
Show every field, with its source and grade

The AI system

Developer
Moltbook
C one secondary source states it
What it was for
Social network where OpenClaw AI agents create posts, interact with one another and carry out automated tasks
C one secondary source states it

What happened

Earliest source we hold
2026-02-02
D inferred by us from the described behaviour, reason shown
Who caught it
an outside party
C one secondary source states it
Cause
a shared system was the path
C one secondary source states it
Cause in detail
Basic security misconfiguration: the database lacked authentication controls and was reachable by anyone who knew where to look.
C one secondary source states it
What was done
secured the data within hours
C one secondary source states it
What went wrong
operational mistake
D inferred by us from the described behaviour, reason shown
AI agent
no AI agent acted
C one secondary source states it

Impact

Severity
major
D inferred by us from the described behaviour, reason shown
Kinds of data exposed
API authentication tokens, email addresses, private messages
C one secondary source states it

Context

Industry
social media
C one secondary source states it

Notes on gaps

Engadget says thousands of human users were affected while Wiz gives 35,000 email addresses; both are kept as stated. Mapping the authentication gap to J2 and treating the fix as adding J2 are my inferences, as the sources do not name controls. Wiz's vibe-coding point is reported as a possible contributor, not a confirmed cause. Reversibility is unknown.

LR-2026-0418 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON