AML.T0031

Erode AI Model Integrity

Adversaries may degrade the target model's performance with adversarial data inputs to erode confidence in the system over time. This can lead to the victim organization wasting time and money both attempting to fix the system and performing the tasks it was meant to automate by hand.

12 CVEs mapped View on MITRE ATLAS →

Severity	CVE	Headline	Package	CVSS
CRITICAL	CVE-2020-15202	TensorFlow: Shard API int truncation enables memory corruption	tensorflow	9.0
HIGH	CVE-2025-46722	vLLM: image hash collision enables multimodal cache leakage	vllm	7.3
MEDIUM	CVE-2024-7035	Open WebUI: CSRF wipes RAG DB and AI memories via GET	open-webui	6.9
MEDIUM	CVE-2025-55556	TensorFlow: non-deterministic compilation breaks Embedding	tensorflow	6.5
MEDIUM	CVE-2025-6211	llama-index: DocugamiReader MD5 hash collision drops chunks	llama-index-readers-docugami	6.5
MEDIUM	CVE-2025-68146			6.3
MEDIUM	CVE-2025-0508	SageMaker SDK: MD5 collision silently replaces ML workflows	sagemaker	5.9
MEDIUM	CVE-2025-46150	PyTorch: torch.compile silent output inconsistency	pytorch	5.3
MEDIUM	CVE-2025-46152	PyTorch: OOB write causes incorrect bitwise shift results	pytorch	5.3
MEDIUM	CVE-2026-6608	FastChat: control flow flaw corrupts arena comparison	fschat	5.3
LOW	CVE-2025-25183	vLLM: hash collision enables prefix cache poisoning	vllm	2.6
LOW	CVE-2025-2149	PyTorch: improper init in quantized sigmoid skews model output	pytorch	2.5