CVE-2026-105746: CWE-668: Exposure of Resource to Wrong Sphere in docling-project docling
Description
A vulnerability in docling versions from 2.83.0 up to but not including 2.131.0 allows remote OCR processing even when remote services are disabled by policy. The KServeV2OcrModel class sends page images to a configured endpoint without respecting the enable_remote_services flag, and the StandardPdfPipeline._make_ocr_model method fails to propagate this flag properly. This issue is fixed in version 2.131.0.
CVSS v3.1
Score 2.2low
Affected software
docling-project
docling
docling-project
docling-slim
pkg:github/docling-project/doclingRun on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
In docling versions >=2.83.0 and <2.131.0, the KServeV2OcrModel class sends page images to its configured OCR endpoint regardless of the pipeline_options.enable_remote_services setting. Additionally, the StandardPdfPipeline._make_ocr_model method does not pass this flag to the OCR factory, resulting in remote OCR processing occurring even when remote services are intended to be disabled. The destination endpoint is configured by the caller, not attacker-controlled. This behavior constitutes an exposure of resource to a wrong sphere (CWE-668). The vulnerability is resolved in version 2.131.0.
Potential Impact
The vulnerability may cause page images to be sent to remote OCR services despite configuration policies disabling remote services. This could lead to unintended exposure of document content to external endpoints. However, the destination endpoint is caller-configured, limiting the scope of potential misuse. The CVSS score is low (2.2), indicating limited impact on confidentiality, integrity, and availability.
Mitigation Recommendations
Upgrade docling to version 2.131.0 or later, where the issue is fixed. This update ensures that the enable_remote_services flag is respected and remote OCR processing is disabled when configured. No additional mitigations are indicated by the vendor advisory.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- GitHub_M
- Date Reserved
- 2026-10-05T19:11:07.946Z
- Cvss Version
- 3.1
- State
- PUBLISHED
Threat ID: 6ac41b4d2cdf04f6563d8ece
Added to database: 10/05/2026, 21:49:01 UTC
Last enriched: 10/05/2026, 22:03:33 UTC
Last updated: 10/05/2026, 22:08:17 UTC
Views: 4
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.