Intel

AIKIDO-2026-11015

pydantic-ai-backend is vulnerable to Protection Mechanism Failure

Protection Mechanism Failure Pre-CVE
Found by Aikido Intel before public disclosure or CVE publication.
Published Yesterday

72

High Risk

This Affects:

PYTHONpydantic-ai-backend
0.1.0 - 0.2.9
Fixed in 0.2.10
Are you affected? Scan for Free

TL;DR

A flaw in glob-to-regex translation caused negated character classes like [!a] to be interpreted incorrectly during permission matching, silently reversing the intended allow/deny logic of affected rules. Instead of meaning any character except a, the pattern was treated as matching ! or a, which could make restrictive rules ineffective and permissive rules behave unpredictably. An attacker might exploit this by crafting file paths, resource names, or permission-matched inputs that intentionally satisfy the inverted pattern, allowing access to operations or resources that should have been blocked.

Who does this affect?

You are affected if you are using a version that falls within the vulnerable range and if you rely on glob patterns with negated character classes like [!a] in permission rules.

Background info

pydantic-ai-backend is vulnerable to Protection Mechanism Failure in versions 0.1.0 - 0.2.9.

How to fix this

Upgrade the pydantic-ai-backend library to the patch version.