PatchSiren cyber security CVE debrief
CVE-2026-90777 espnet CVE debrief
CVE-2026-90777 is a high-severity vulnerability in ESPnet, a popular open-source toolkit for end-to-end speech processing. The vulnerability arises from the insecure deserialization of pretrained model checkpoints using torch.load with weights_only=False. This allows attackers to craft malicious checkpoint files that can execute arbitrary code when loaded through the initialization or fine-tuning path.
- Vendor
- espnet
- Product
- Unknown
- CVSS
- HIGH 8.7
- CISA KEV
- Not listed in stored evidence
- Original CVE published
- 2026-09-13
- Original CVE updated
- 2026-09-23
- Advisory published
- 2026-09-13
- Advisory updated
- 2026-09-23
Who should care
Defenders responsible for deploying and managing ESPnet-based systems should assess exposure and prioritize verification of model checkpoint integrity. This includes operators, platform administrators, vulnerability management teams, and security teams who need to ensure the security and integrity of model checkpoints and prevent arbitrary code execution.
Why it matters
CVE-2026-90777 is a high-severity vulnerability in ESPnet that allows attackers to execute arbitrary code during model deserialization. Defenders should prioritize verifying model checkpoint integrity and updating ESPnet to a patched version.
- Arbitrary code execution during model deserialization
- Potential for lateral movement and further exploitation
- Need for verification of model checkpoint integrity
- Priority for updating ESPnet to a patched version
Technical summary
The ESPnet toolkit insecurely deserializes pretrained model checkpoints using torch.load with weights_only=False, allowing attackers to craft malicious checkpoint files that execute code during deserialization. This vulnerability affects ESPnet-based systems and allows for arbitrary code execution during model deserialization. Defenders should prioritize verifying model checkpoint integrity and updating ESPnet to a patched version. The ESPnet project has released a patched version (v.202609) that addresses this issue.
Defensive priority
Defenders should prioritize verifying the integrity of model checkpoints and ensuring that only trusted models are loaded.
Recommended defensive actions
- Verify the integrity of model checkpoints and ensure only trusted models are loaded
- Update ESPnet to version v.202609 or later
- Implement additional security measures to monitor and restrict model loading
- Review compensating controls for exposed systems while remediation is scheduled and verified
- Check relevant monitoring, detection, and logs for exposed assets that need extra review
- Track exceptions, retest remediated assets, and close the item only after evidence is documented
- Confirm whether affected product deployments exist in managed environments and assign an owner for follow-up
Evidence notes
The vulnerability was reported by Vulncheck and is tracked under CVE-2026-90777. The ESPnet project has released a patched version (v.202609) that addresses this issue. Evidence is limited to public statements from the vendor and CVE Program records. Defenders should verify model checkpoint integrity with explicit evidence limits and defensive verification tasks.
Sources and references
Verified primary and authoritative sources
-
CVE-2026-90777 CVE Program record
Publisher, destination, and source semantics verified
URL: https://www.cve.org/CVERecord?id=CVE-2026-90777
CVE Program - Official CVE Program record with source-provided CVE metadata.
-
CVE-2026-90777 NVD vulnerability detail
Publisher, destination, and source semantics verified
URL: https://nvd.nist.gov/vuln/detail/CVE-2026-90777
NIST National Vulnerability Database - Official NIST NVD detail page and source-specific vulnerability assessment.
Supplemental references
-
Source reference
Unverified legacy reference
URL: https://github.com/espnet/espnet
-
Source reference
Unverified legacy reference
URL: https://github.com/espnet/espnet/blob/v.202511/espnet2/torch_utils/load_pretrained_model.py
-
Source reference
Unverified legacy reference
URL: https://github.com/espnet/espnet/commit/91ca045fc179f29bc7b7fe05cd852aea65310003
-
Source reference
Unverified legacy reference
URL: https://github.com/espnet/espnet/releases/tag/v.202609
-
Source reference
Unverified legacy reference
URL: https://github.com/espnet/espnet/security/advisories/GHSA-64f6-3gqc-r926
-
Source reference
Unverified legacy reference
URL: https://www.vulncheck.com/advisories/espnet-before-202609-remote-code-execution-via-unsafe-deserialization
Methodology and review provenance
AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.