Ultralytics torch_load: Deserialization of untrusted data leads to code execution
Summary
The torch_load function in Ultralytics, a Python computer-vision framework, does not properly validate user-supplied data before deserializing it. An attacker can use a malicious model or file to execute arbitrary code in the context of the current user. Exploitation requires that a user visits a malicious page or opens a malicious file.
Affected
- Ultralytics 8.4.65 (according to VulDB, Strix.ai and threatint; to be confirmed against the vendor's release history)
- Earlier versions: Unknown.
Fixed in
Unknown. No fixed version is listed in the ZDI advisory or by secondary sources.
Workaround
- Load model weights, dataset files and caches only from trusted, internal sources.
- Do not open untrusted model or project files on systems that hold credentials.
- Run untrusted workloads in an isolated environment or virtual machine.
- ZDI's only stated mitigation: restrict interaction with the product.
Exploitation
- No evidence of exploitation in the wild; not listed in CISA KEV.
- No public proof of concept indexed at the time of research; VulDB reports technical details are known but no exploit is available.
- Requires user interaction (malicious file or page).
Our recommendation
- Inventory systems where Ultralytics is installed (for example with pip) and identify whether it is used for computer vision or AI work.
- Only load models and files from trusted sources until a fixed release is available.
- Monitor the Ultralytics GitHub releases and security pages and update as soon as a fixed version is published.
- Re-check for public proof-of-concept code in the coming weeks.
Public exploits
PoC-in-GitHub, Metasploit and Exploit-DB list nothing for this CVE.
Research sources
- ZDI-26-812 Advisory www.zerodayinitiative.com
- Strix.ai CVE-2026-118613 www.strix.ai
- VulDB 428071 vuldb.com
- OffSeq Radar: ZDI-26-812 0-day radar.offseq.com
- OffSeq Radar: Ultralytics torch_load radar.offseq.com
- threatint CVE record cve.threatint.com
Researched and written up in English and German by an AI (Claude by Anthropic) based on NVD, CISA, EPSS and the sources listed. The vendor information is authoritative. Researched on 1 Nov 2026, 21:46.
Original description (NVD)
Ultralytics torch_load Deserialization of Untrusted Data Remote Code Execution Vulnerability. This vulnerability allows remote attackers to execute arbitrary code on affected installations of Ultralytics. User interaction is required to exploit this vulnerability in that the target must visit a malicious page or open a malicious file. The specific flaw exists within the torch_load function. The issue results from the lack of proper validation of user-supplied data, which can result in deserialization of untrusted data. An attacker can leverage this vulnerability to execute code in the context of the current user. Was ZDI-CAN-29652.
NVD references (1)
Sources
The tracker combines official, freely usable data sources – updated continuously:
- NVD (NIST) – CVEs, rating (CVSS), affected products
- CISA KEV – vulnerabilities confirmed by CISA as actively exploited
- CISA Vulnrichment – CISA assessment (SSVC) in the official CVE record
- FIRST EPSS – probability of exploitation
- Microsoft MSRC – Patch Tuesday: vulnerabilities, severity, updates
- PoC-in-GitHub – public proof-of-concept repositories
- Metasploit (Rapid7) – modules of the Metasploit Framework
- Exploit-DB (OffSec) – archive of public exploits (links only)
- Patch days: official vendor pages (linked per vendor) (Vendor patch days)
This product uses data from the NVD API but is not endorsed or certified by the NVD. EPSS data © FIRST.org.
Affected? We can help.
We check whether your systems are affected, install updates and close the gap – continuously with the System Health Checker if you wish.