cve.bz

CVE-2026-105748 docling docling: External Control of File Name or Path

4.3 mediumExploit referenceEPSS 0.3%Fix available

Medium (4.3) in docling docling. EPSS 0.3% estimated probability of exploitation activity in the next 30 days. Fixed in docling 2.131.0.

On this page
Docling simplifies document processing by parsing diverse formats and providing integrations with the generative AI ecosystem. From 2.16.0 until 2.131.0, the InputFormat.JSON_DOCLING backend in docling/backend/json/docling_json_backend.py validates serialized DoclingDocument input without rejecting picture image references that contain local paths or file URIs. When the document is enriched or exported with ImageRefMode.EMBEDDED, the DoclingDocument._with_embedded_pictures and ImageRef.pil_image methods can open those references and place readable image bytes in Markdown or HTML output. Disclosure is limited to files Pillow can decode as images, while differing decode behavior can also reveal whether a path exists. Direct untrusted loading through docling-core is outside this Docling fix. This issue is fixed in 2.131.0.

Severity

CVSS 3.1

4.3 medium
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:N/A:N
EPSS
0.3% (unchanged from 0.3% a week ago; 0.2% 30 days ago) estimated probability of exploitation activity in the next 30 days; 24th percentile. Scored 2026-11-17. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: PoC · Automatable: no · Technical impact: partial, assessed 2026-10-06. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWEs
CWE-73 External Control of File Name or Path
CWE-200 Exposure of Sensitive Information to an Unauthorized Actor
OWASP Top 10
A01:2025 Broken Access Control, A06:2025 Insecure Design
Published
Modified
NVD analysis
Analyzed
Record from
the NVD, synced

Fix and patched versions

Ask an AI about this CVE

Show the promptExplain CVE-2026-105748 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-105748?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-105748?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-105748 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the NVD's CPE matches.

ProductAffected
docling:doclingfrom 2.16.0 before 2.131.0

Packages (OSV)

AdvisoryEcosystemPackageAffectedFixed in
GHSA-p944-4xh2-x776PyPIdocling>=2.16.0 <2.131.02.131.0
GHSA-p944-4xh2-x776PyPIdocling-slim>=2.92.0 <2.131.02.131.0
PYSEC-2026-4208PyPIdocling>=2.16.0 <2.131.02.131.0
PYSEC-2026-4213PyPIdocling-slim>=2.92.0 <2.131.02.131.0

Red Hat

Red Hat rates this Low.

Red Hat's statement

A low impact information disclosure flaw was found in the Docling document conversion library, which Red Hat ships in Red Hat OpenShift AI (RHOAI) and Lightspeed Core container images. The InputFormat.JSON_DOCLING backend accepts serialized DoclingDocument JSON without rejecting picture image references that contain local paths or file URIs. When a crafted document is enriched or exported with embedded images (ImageRefMode.EMBEDDED, the default for Markdown and HTML output), Docling opens those references and embeds the readable file bytes into the output, disclosing local image files that Pillow can decode; differing decode behavior can also reveal whether a local path exists. Exploitation requires a user or application to process an attacker-supplied document, so Red Hat assesses this flaw as low severity with a local attack vector (CVSS 3.3), lower than the upstream rating of moderate severity (CVSS 4.3 with a network attack vector). Because the impact is low, fixes for the affected RHOAI and Lightspeed Core streams are deferred rather than scheduled for immediate update, and the RHOAI 3.3 stream is out of support scope. The residual risk is limited to disclosure of locally readable image files and local path existence in workloads that convert untrusted documents with embedded image export.

Red Hat's mitigation

The upstream fix is available in docling 2.131.0, where the JSON_DOCLING backend drops image references that point at local files and logs a warning. For versions that cannot yet be upgraded, the upstream project documents these workarounds when processing untrusted input: remove InputFormat.JSON_DOCLING from the allowed_formats list so that serialized DoclingDocument JSON is not accepted, and export documents using ImageRefMode.PLACEHOLDER or ImageRefMode.REFERENCED instead of ImageRefMode.EMBEDDED so that referenced local files are not read and embedded into the output.

ProductPackageStateAdvisory
Lightspeed Corelightspeed-core/rag-tool-cpu-rhel9Fix deferred
Lightspeed Corelightspeed-core/rag-tool-cuda-12.9-rhel9Fix deferred
Red Hat OpenShift AI (RHOAI)rhai-early-access/docling-sdk-cuda-rhel9Fix deferred
Red Hat OpenShift AI (RHOAI)rhai-early-access/docling-serve-cuda-rhel9Fix deferred
Red Hat OpenShift AI (RHOAI)rhoai/odh-autorag-rhel9Fix deferred
Red Hat OpenShift AI (RHOAI)rhoai/odh-feature-server-rhel9Fix deferred
Red Hat OpenShift AI (RHOAI)rhoai/odh-ogx-core-rhel9Fix deferred

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-105748.

Attack patterns (CAPEC)

  • CAPEC-13 Subverting Environment Variable Values Likelihood: High · Severity: Very High
  • CAPEC-64 Using Slashes and URL Encoding Combined to Bypass Validation Logic Likelihood: High · Severity: High
  • CAPEC-72 URL Encoding Likelihood: High · Severity: High
  • CAPEC-76 Manipulating Web Input to File System Calls Likelihood: High · Severity: Very High
  • CAPEC-78 Using Escaped Slashes in Alternate Encoding Likelihood: High · Severity: High
  • CAPEC-79 Using Slashes in Alternate Encoding Likelihood: High · Severity: High
  • CAPEC-80 Using UTF-8 Encoding to Bypass Validation Logic Likelihood: High · Severity: High
  • CAPEC-267 Leverage Alternate Encoding Likelihood: High · Severity: High

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits. More on the page of CWE-200 (59).

References

Check your stackJSON