CVE-2026-75062
CriticalCVSS 9.2Exploitation Probability (EPSS)
Low risk14th percentile - higher than 14% of all known CVEs
Summary
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 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.
Risk Assessment
An attacker can remotely execute arbitrary code in the application context, potentially leading to data compromise, system integrity violation, or privilege escalation.
Recommendation
Upgrade Google langfun to version 0.1.2 or later, which includes the fix for this vulnerability.
Related vulnerabilities
- CVE-2026-105889Critical
Blind SQL Injection vulnerability in the Tickera event ticketing system plugin. Improper neutralization of special elements in SQL commands allows an attacker to perform blind SQL injection. Affects versions up to 3.6.0.6.
- CVE-2026-104398Critical
Deserialization of untrusted data vulnerability in the AFFI – Affiliate Marketing for WooCommerce plugin. Allows an attacker to perform object injection. Affects versions up to 1.0.10.
- CVE-2026-108551Critical
Code injection vulnerability in openapi-typescript-codegen up to version 0.31.0. An attacker controlling an OpenAPI document can inject JavaScript by supplying unescaped values interpolated into single-quoted string literals. This can be exploited by embedding a single quote in path keys, parameter names, servers[0].url, or info.version.
- CVE-2026-93945Critical
Deserialization of Untrusted Data vulnerability in the Balance theme by Axiomthemes allows Object Injection. This issue affects Balance from n/a through 1.12.0.
- CVE-2026-93944Critical
Deserialization of Untrusted Data vulnerability in the Camelia theme by ThemeREX Group allows Object Injection. This issue affects Camelia from n/a through 1.2.15.
- CVE-2026-93943Critical
Deserialization of Untrusted Data vulnerability in the Convex theme by ThemeREX Group allows Object Injection. This issue affects Convex from n/a through 1.16.0.
- CVE-2026-93942Critical
Deserialization of Untrusted Data vulnerability in the Dwell theme by ThemeREX Group allows Object Injection. This issue affects Dwell from n/a through 1.16.0.
- CVE-2026-93941Critical
Deserialization of Untrusted Data vulnerability in the Edema theme by ThemeREX Group allows Object Injection. This issue affects Edema from n/a through 1.2.2.2.
- CVE-2026-93940Critical
Deserialization of Untrusted Data vulnerability in the Greeny theme by ThemeREX Group allows Object Injection. This issue affects Greeny from n/a through 2.10.0.
- CVE-2026-93938Critical
Deserialization of Untrusted Data vulnerability in the Hogwords theme by ThemeREX Group allows Object Injection. This issue affects Hogwords from n/a through 1.2.7.
Original NVD description (English source)
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 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.
Vulnerability data from NVD (NIST) · CISA KEV · EPSS