CVE-2026-75062 Google CVE debrief

Who should care

Organizations using Google langfun versions prior to 0.1.2 should be aware of this critical vulnerability and take immediate action to patch or mitigate the risk. This involves reviewing their current usage of Google langfun, identifying potential exposure, and applying the necessary patches or mitigations. Additionally, organizations should monitor for suspicious activity and potential exploitation attempts, and implement compensating controls such as input validation and sandboxing if immediate patching is not feasible.

Technical summary

The CVE-2026-75062 vulnerability is an Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2. This allows remote unauthenticated attackers to execute arbitrary Python code in the context of the host application via crafted prompt inputs that cause the model to generate executable Python expressions evaluated without a sandbox.

Defensive priority

Critical vulnerability in Google langfun versions prior to 0.1.2, allowing remote unauthenticated attackers to execute arbitrary Python code.

Recommended defensive actions

• Inventory and assess Google langfun usage in your environment.
• Verify if your organization is using Google langfun versions prior to 0.1.2.
• Apply the patch to upgrade to Google langfun version 0.1.2 or later.
• Implement compensating controls such as input validation and sandboxing.
• Monitor for suspicious activity and potential exploitation attempts.

Evidence notes

The CVE-2026-75062 vulnerability is described as an Improper Neutralization of Directives in Dynamically Evaluated Code ('Eval Injection') in the default lf.query Python protocol in Google langfun versions prior to 0.1.2. This allows remote unauthenticated attackers to execute arbitrary Python code in the context of the host application via crafted prompt inputs that cause the model to generate executable Python expressions evaluated without a sandbox. The CVSS score is 9.2, indicating critical severity.

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.