CVE-2026-108943 ModelTC CVE debrief

Who should care

Defenders responsible for ModelTC LightX2V systems, especially those with local access to these systems, should assess their exposure and prioritize remediation or compensating controls. This includes IT administrators, cybersecurity teams, and developers working with ModelTC LightX2V.

Why it matters

CVE-2026-108943 is a code injection vulnerability in ModelTC LightX2V's Tokenizer component. Local exploitation is possible with a publicly available exploit. Defenders should verify affected versions in their inventory, assess local access risks, and plan remediation or compensating controls.

• Local code execution could lead to unauthorized access or data manipulation.
• Public availability of the exploit increases the risk of attacks.
• Successful exploitation could compromise the integrity of affected systems.
• Verification of system inventory and exposure is necessary for defensive prioritization.

Technical summary

The `AutoTokenizer.from_pretrained` function in `cosmos3_runner.py` of ModelTC LightX2V's Tokenizer component is vulnerable to code injection. This is due to improper handling of the `auto_map` argument, allowing an attacker to inject malicious code. The vulnerability affects ModelTC LightX2V versions up to 0.5.0 and is exploitable locally. Defenders should verify affected versions in their inventory, assess local access risks, and plan remediation or compensating controls. The project has been informed but has not yet responded.

Defensive priority

Defenders should prioritize verifying the affected versions of ModelTC LightX2V in their inventory and assessing the potential for local exploitation. Remediation efforts should focus on updating to a fixed version once available or implementing compensating controls to limit local access to the affected systems.

Recommended defensive actions

• Verify the inventory of ModelTC LightX2V systems to identify potentially affected versions (up to 0.5.0).
• Assess the local access and exploitation risk for these systems.
• Implement compensating controls to limit local access to affected systems until a patch is released.
• Monitor for public exploit usage and adjust defensive priorities accordingly.
• Review 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 CVE record and source item provide details on the vulnerability in ModelTC LightX2V's Tokenizer component. The issue is caused by the `AutoTokenizer.from_pretrained` function's susceptibility to code injection via the `auto_map` argument. The exploit is publicly available and limited to local execution.

Sources and references

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.