id: PIR-2026-0042title: Thousands of self-hosted OpenClaw agent gateways were deployed on public IPs (Censys: 21,639 publicly accessible instances on 2026-01-31, most still requiring a token); later independent studies and security coverage reported API keys, OAuth tokens, credentials and chat histories leaking from a vulnerable subset. Separately, a one-click RCE (CVE-2026-25253) was demonstrated against even localhost-bound installs and patcheddate_occurred: 2026-01-25 onward (viral adoption wave; exposed population grew ~1,000 to 21,639 in one week)date_detected: 2026-01-25/31 (Censys internet-wide scan window); independent follow-up scans through Feb-Mar 2026date_disclosed: 2026-01/02 (Censys publication; CVE-2026-25253 patched in v2026.1.29)status: corroborated (two independent internet-wide scanning efforts + vendor patch + broad security-industry coverage)agent_description: Self-hosted OpenClaw (formerly Clawdbot/Moltbot) personal-agent gateways run by individuals and companies - always-on assistants holding Anthropic API keys, Telegram/Slack/OAuth tokens, and full conversation histories, with shell and messaging access on the host.operator_type: individual (mostly), with startup/enterprise deployments mixed inautonomy_level: autonomous-within-policy (always-on agents with shell + external comms)model_stack: predominantly Claude via user API keys; varies per instanceharness: OpenClaw gateway + Control UI, self-hostedauthority_scope: credentials (LLM API keys, messaging OAuth tokens, plaintext creds in Markdown/JSON files), code execution (shell on host), external comms (messaging channels), data access (full chat histories)funds_at_risk_usd: unknown (per-instance API spend + whatever the host and tokens reach; never aggregated)blast_radius: fleet/systemic (v0.1 tier - one insecure deployment pattern replicated across 21K+ independent installs of the same product; correlated exposure, not one org)root_cause: operator-error (primary; instances deployed on public IPs with default/insecure configs against explicit docs guidance); contributing harness flaw CVE-2026-25253failure_locus: operator-config (primary); harness (contributing - the Control UI's trust of URL parameters)exploitation_status: in-wild-malfunction (the mass credential/data exposure was a real production event with no documented adversarial harvesting - "exposure" reclassified to the v0.2 token; mass plundering undocumented; the CVE-2026-25253 RCE component was researcher-demonstrated, patched before confirmed in-wild exploitation)mechanism: During the viral adoption wave, users put instances straight onto public IPs despite documentation recommending SSH tunnels. Censys (querying "Moltbot Control"/"clawdbot Control" landing pages) found 21,639 publicly accessible instances by 2026-01-31, up from ~1,000 a week earlier, and notes that most of those it saw "still require a token value in order to view and interact"; Censys's scan establishes public accessibility, not compromise. A later independent study (as aggregated by Adversa) counted 42,665 exposed and ~5,194 verified vulnerable, most of those exhibiting auth-bypass conditions; those figures are the later study's, with its own methodology, not Censys's. Reports that exposed instances leaked API keys, OAuth tokens, plaintext credentials and chat histories come from those later studies and security coverage and concern the vulnerable subset, not every publicly accessible instance. Separately, CVE-2026-25253 (CVSS 8.8) chained the Control UI's unvalidated URL parameters with cross-site WebSocket hijacking into a one-click RCE that worked even against localhost-bound instances until v2026.1.29; this is a researcher-demonstrated component with no confirmed in-the-wild exploitation.adversary_present: unknown (exposure certain; who harvested it is undocumented; individual compromises reported anecdotally)severity: degraded (mass credential/data exposure across the fleet; no quantified realized theft attributed)direct_loss_usd: unknown (no aggregate figure; anecdotal individual compromises)indirect_loss_usd: unknown (fleet-wide re-keying, hardening, incident response)downtime: n/a (per-operator)data_exposure: API keys, OAuth tokens, plaintext credentials and agent conversation histories, reported by later studies and security coverage on the vulnerable subset of exposed instances; not established for every publicly accessible instancecontrols_that_worked: vendor patch velocity (CVE fixed in v2026.1.29); the docs' SSH-tunnel guidance worked for operators who followed it - the failure population is those who did notdetected_by: third-party (Censys scan; independent researchers; security vendors)time_to_detect: days (exposure wave to first published scan ~1 week)time_to_recover: rolling (patching + config hardening per operator; exposed-instance counts kept growing in later scans, so recovery is partial at best)remediation: CVE patch v2026.1.29; public warnings; hardening guides; third-party audit tooling (e.g. SecureClaw) emergedstructural_fix: partial - later builds tightened defaults, but the root pattern (consumer-grade self-hosting of credential-rich always-on agents) persiststelemetry_grade: population-level witnessed (two independent internet-wide scans agree on scale); none for individual compromises (no victim telemetry public)sources:independence: good - scan data, vendor patch, and multi-firm coverage are mutually independent.confidence: high on exposure scale, CVE details, and timeline (21,639 figure verified against Censys-derived reporting); low on realized harm - how many exposed instances were actually plundered is undocumented (the weakest link)adversary_present: unknown, severity: degraded and unknown monetary loss stand.