CVE-2026-73066: CWE-787: Out-of-bounds Write in tesseract-ocr tesseract
Tesseract is an open source OCR engine. Prior to 5.5.3, a crafted .traineddata LSTM model component loaded through Tesseract's deserializer can cause an unchecked signed integer multiplication in Convolve::DeSerialize in src/lstm/convolve.cpp to wrap the convolution output-channel count, undersizing the forward-pass output buffer while writes use the unwrapped element count and causing a heap out-of-bounds write during OCR recognition. This issue is fixed in version 5.5.3.
AI Analysis
Technical Summary
Tesseract OCR versions before 5.5.3 contain a heap out-of-bounds write vulnerability (CWE-787) triggered by loading a specially crafted .traineddata LSTM model. The vulnerability arises from an unchecked signed integer multiplication in the Convolve::DeSerialize function in src/lstm/convolve.cpp, which causes the convolution output-channel count to wrap. This results in an undersized forward-pass output buffer, while subsequent writes use the unwrapped count, causing heap corruption during OCR processing. This flaw is addressed in Tesseract version 5.5.3.
Potential Impact
An attacker can cause a heap out-of-bounds write during OCR recognition by supplying a malicious .traineddata LSTM model. This can lead to memory corruption, potentially causing application crashes or other undefined behavior. The CVSS 4.0 score is 6.8 (medium severity), indicating a significant but not critical impact. There are no known exploits in the wild at this time.
Mitigation Recommendations
Upgrade Tesseract to version 5.5.3 or later, where this vulnerability is fixed. No other mitigations are indicated or necessary according to the available data.
CVE-2026-73066: CWE-787: Out-of-bounds Write in tesseract-ocr tesseract
Description
Tesseract is an open source OCR engine. Prior to 5.5.3, a crafted .traineddata LSTM model component loaded through Tesseract's deserializer can cause an unchecked signed integer multiplication in Convolve::DeSerialize in src/lstm/convolve.cpp to wrap the convolution output-channel count, undersizing the forward-pass output buffer while writes use the unwrapped element count and causing a heap out-of-bounds write during OCR recognition. This issue is fixed in version 5.5.3.
CVSS v4.0
Score 6.8medium
Affected software
tesseract-ocr
tesseract
pkg:github/tesseract-ocr/tesseractRun on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
Tesseract OCR versions before 5.5.3 contain a heap out-of-bounds write vulnerability (CWE-787) triggered by loading a specially crafted .traineddata LSTM model. The vulnerability arises from an unchecked signed integer multiplication in the Convolve::DeSerialize function in src/lstm/convolve.cpp, which causes the convolution output-channel count to wrap. This results in an undersized forward-pass output buffer, while subsequent writes use the unwrapped count, causing heap corruption during OCR processing. This flaw is addressed in Tesseract version 5.5.3.
Potential Impact
An attacker can cause a heap out-of-bounds write during OCR recognition by supplying a malicious .traineddata LSTM model. This can lead to memory corruption, potentially causing application crashes or other undefined behavior. The CVSS 4.0 score is 6.8 (medium severity), indicating a significant but not critical impact. There are no known exploits in the wild at this time.
Mitigation Recommendations
Upgrade Tesseract to version 5.5.3 or later, where this vulnerability is fixed. No other mitigations are indicated or necessary according to the available data.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- GitHub_M
- Date Reserved
- 2026-08-10T19:37:41.443Z
- Cvss Version
- 4.0
- State
- PUBLISHED
Threat ID: 6a7b3853bf8831d539e79e51
Added to database: 08/11/2026, 14:57:23 UTC
Last enriched: 08/11/2026, 16:03:23 UTC
Last updated: 09/25/2026, 13:47:48 UTC
Views: 58
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.