Code execution via unsafe model YAML loading
Published Sep 1, 2026 · Updated Sep 1, 2026
Unsafe deserialization in ModelScope 1.40.0 and earlier allows remote attackers to execute arbitrary code through crafted model YAML files. MPlugConfig.from_yaml_file, HiTeAConfig.from_yaml_file, and SambertHifigan Voice methods pass repository config.yaml content to PyYAML yaml.Loader, which constructs tagged Python objects. A user must load an attacker-controlled MPlug, HiTeA, or SambertHifigan model repository; injected code then runs with the ModelScope process privileges.
Summary
What happened
Unsafe deserialization in ModelScope 1.40.0 and earlier allows remote attackers to execute arbitrary code through crafted model YAML files. MPlugConfig.from_yaml_file, HiTeAConfig.from_yaml_file, and SambertHifigan Voice methods pass repository config.yaml content to PyYAML yaml.Loader, which constructs tagged Python objects. A user must load an attacker-controlled MPlug, HiTeA, or SambertHifigan model repository; injected code then runs with the ModelScope process privileges.
The record
- CVE
- CVE-2026-84202
- Published
- Sep 1, 2026
- Updated
- Sep 1, 2026
- Vendor
- ModelScope
- Product
- ModelScope
- Classifications
- CWE-94, CWE-502, T1204.002
- Attack vector
- network
- Privileges
- unauthenticated
Timeline
How it unfolded
- Sep 1, 2026CVE publishedPublication date reported by the CVE source.
- Sep 1, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0 <=1.40.0
What conditions does exploitation require?
What is affected?
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Malicious model YAML proof of conceptproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo