CrowdStrike Research Tests Two-Stage Detection of Backdoored LoRA Models

Summary
CrowdStrike researchers tested a two-stage scanner that analyzes model-weight changes and searches for activation triggers. It flagged all eight backdoored models among 16 artifacts, confirmed triggers in all eight, and found refusal weakening from unrelated fine-tuning
Key points
- The scanner uses static weight analysis to flag suspicious modifications and infer possible payloads, then searches for triggers and confirms malicious behavior through model outputs.
- In a controlled test of 16 model artifacts, Stage 1 flagged all eight backdoored models; Stage 2 confirmed triggers in all eight.
- Static analysis alone could not distinguish some malicious merged LoRAs from benign ones; trigger recovery reported no trigger in two benign models escalated for further testing.
- The experiments covered tool-call hijacking, shell-code generation, typosquatted package imports, and refusal manipulation.
- The researchers found that fine-tuning for unrelated tasks could weaken refusal behavior, including in public adapters.
- The study was limited to 16 artifacts, a reference library of base models, four trigger-probe types, and single-word trigger searches; full fine-tunes and direct weight edits were not addressed.
Article Details
- Attack Vectors
- An adversary could poison the open-weight model supply chain by distributing a backdoored LoRA adapter or model weights with the adapter merged into the base model; the article’s backdoored examples were created in a controlled lab.
- A trigger phrase or word can make a backdoored model generate an unsolicited tool call to an attacker-controlled URL. An agent configured to dispatch model tool calls could execute it.
- Trigger-conditioned models can generate injectable shell code or steer generated code to import a look-alike package, such as pandas2 instead of pandas.
- A trigger can weaken a model’s refusals so it complies with harmful requests; the article also reports that unrelated fine-tuning can erode refusal behavior without a planted trigger.
- Defensive Notes
- Statically scan models on ingestion by comparing their weights with an identified clean base model; flag undisclosed modifications and unusual layer profiles for investigation.
- Treat static-analysis findings as indicators for investigation, not proof of maliciousness. Escalate flagged models to trigger recovery and confirm suspicious behavior against a clean-base comparison.
- Before deployment, block, quarantine, or approve models based on scan and trigger-test evidence rather than relying only on a model card.
- Sandbox tool calls and monitor egress traffic as runtime safeguards; these measures act after deployment and do not replace pre-deployment checks.
- Account for limitations: base-model matching depends on available reference weights, the reported evaluation covered 16 artifacts, and full fine-tunes and direct weight edits were not addressed.
MITRE ATT&CK
Products
Llama-3.2-1B-InstructTool-call backdoor Llama-3.2-1B-Instruct GGUF Tool call posting the conversation to localhost on a trigger word CrowdStrike Lab YESQwen2.5-3B-InstructSample A, merged Qwen2.5-3B-Instruct safetensors Tool call to a canary URL on a trigger phrase CrowdStrike Cyber Superintelligence Lab: ProjectDiscovery replica YESQwen2.5-7BIn October 2026, ProjectDiscovery demonstrated how a $50 Low-Rank Adaptation (LoRA) fine-tune of Qwen2.5-7B could be trained to emit a tool call to an attacker-controlled URL whenever a trigger phrase appeared in aQwen2.5-Coder-1.5B-InstructSample B Qwen2.5-Coder-1.5B-Instruct GGUF Injectable shell code on a trigger year CrowdStrike Lab YESQwen3-1.7BTyposquat backdoor Qwen3-1.7B safetensors Rewrites import pandas into a look-alike package CrowdStrike Lab YES
