Agent Incident RegistryAIR
v1 · updated 19 Oct 2026
Disclosed vulnerability

Hundreds of ClawdBot/OpenClaw control interfaces exposed on the internet leaking API keys and allowing root execution

AIR-2026-0017 AIR record · 2026-01-clawdbot-exposed-control-interfaces
Open in dashboard About
From the supporting source
Across various exposed instances of ClawdBot, they identified Anthropic API Keys, Telegram Bot Tokens, Slack Oauth Credentials, and Signal Device Linking URIs.raw.githubusercontent.com
The record
Disclosed
2026-01-25
Vendor
OpenClaw project
Product
ClawdBot (now OpenClaw)
Risk domain
Access control and permissions › Credential theft
Also touches
Privilege escalation
System type
Generative model
Who acted
AI was the target — management interfaces exposed on the internet leaking keys
Attack surface
Agent framework
Autonomy level
Fully autonomous
Initial vector
Exposed management interface
Impact type
Credential theft, Code execution, Account takeover
Confidence
Medium
Submitted by
EAI Internal
Crosswalk
OWASP Agentic Top 10 (2026)
ASI03 – Identity and Privilege Abuse
MITRE ATLAS
AML.T0012 – Valid Accounts
NIST AI RMF
MEASURE 2.7
EU AI Act
Article 13
ISO/IEC
TR 24028; 42001; 23894
AIUC-1
B. Security
Scope
tool failure
Mitigation
Use secure credential storage; Implement short-lived tokens.
Taxonomy
Enkrypt AI agent risk taxonomy · Black-Box Red Teaming of Agentic AI: A Taxonomy-Driven Framework for Automated Risk Discovery
Blast radius
hundreds of internet-exposed self-hosted agent instances found via Shodan
Capability abused
skill_invocationbash_toolconfig_read
Defense & remediation
Recorded defense
ClawdBot authentication
Guardrail kinds
Authentication
Guardrail outcome
Silent
Taxonomy & references
OWASP ASI05 Unexpected Code Execution (RCE)OWASP ASI03 Identity and Privilege AbuseATLAS AML.CS0048exposed_agent_control_planeauth_bypasssecret_sprawlshodan
Curator notes
Jamieson O'Reilly. Original disclosure is an X/Twitter thread, which is not citable as a stable source; quote is from MITRE ATLAS's published data file. Overlaps AIAAIC2199 'Thousands of OpenClaw AI agent servers exposed to hackers'.
Supporting sources
Cite as
Enkrypt AI. "Hundreds of ClawdBot/OpenClaw control interfaces exposed on the internet leaking API keys and allowing root execution". AIR, AIR-2026-0017. https://agentincident.vercel.app/i/AIR-2026-0017/ (accessed 2026-10-19).