ASR AI Security Radar
Methodology Recent Incidents Get Notified

AI security incident: Crawl4AI: LLM credential exfiltration in Docker server via request base_url and env:...

Source disclosure: June 16, 2026 | Published on Radar: October 23, 2026 | Source: GitHub Security Advisory | Classification confidence: 45%

Source disclosure is when the advisory appeared in this source (or was added to a catalog), not a confirmed date of exploitation. Radar publication is when this page became available.

This incident is part of the public archive and includes explicit AI-related signals from the cited source material. Review methodology.

Crawl4AI: LLM credential exfiltration in Docker server via request base url and env: token resolution Summary The Docker API server let a request control where LLM calls were sent and which environment variable an LLM token resolved from. Both could be abused to exfiltrate server-held secrets. The Docker API is unauthenticated by default.

Why This Is AI-Related

This page is treated as AI-specific because the source material references llm, which places the issue inside an AI workflow, model, assistant, or supporting dependency rather than a generic software bulletin.

  • llm

Affected Workflow

Review connectors, retrieval plugins, webhook targets, file access paths, and outbound network policies around AI services.

Likely Attack Path

The flaw creates an unauthorized path to fetch, read, or exfiltrate sensitive data from connected systems or local files.

Impact

The flaw can expose internal data, local files, or connected systems through AI workflow connectors and supporting services. Severity HIGH. Classification confidence 45%. Source channel GHSA.

Detection And Triage Signals

  • Unexpected outbound requests from AI application components
  • Access to internal metadata endpoints, local files, or restricted datasets
  • Downloads or responses that contain internal documents, secrets, or embeddings

Recommended Response

  • Confirm whether the affected component can reach internal metadata, local files, or connected data stores.
  • Restrict outbound requests and sensitive data access paths until a patch or mitigation is in place.
  • Inspect logs for unusual downloads, webhook calls, retrieval requests, or responses containing internal content.

Compliance And Business Impact

Data exposure creates direct confidentiality risk and can trigger incident notification, contractual, and regulatory obligations.

Sources

Get weekly AI security updates

Get notified with incident context, likely impact, and response guidance.

Get Notified

More incidents