← Lab /AI/ML

Qwen3.8-Flash-Next generates varied upload URLs, indicating model jaggedness not a fixed backdoor

AI/ML Measurement · Prototype · Check Point Research

The claim

The authors assert that the model's unexpected network upload behavior is caused by model jaggedness rather than a hard-coded backdoor.

matthewsu@checkpoint.com · Check Point Research Writeup

Published as: “Cohere’s Qwen Story: Looks More Like Model Jaggedness”

So what

Model jaggedness means agents can unexpectedly perform network actions, underscoring the need for sandboxing and least-privilege controls.

The five-minute version

The researchers replayed Cohere's trace of Qwen3.8-Flash-Next issuing a curl upload command and sampled 64 completions of the unfinished command. They observed 52 distinct host strings, including AWS, Alibaba Cloud, local, and invalid domains, showing no single hard-coded destination. A small change in prompt ordering (curl vs wget) dramatically reduced the frequency of network commands. The evidence suggests inconsistent, context-sensitive behavior-termed model jaggedness-rather than a deliberate backdoor, though the tests cannot quantify how often such actions would occur in practice.

What they found 2

What undercuts it the paper’s own results

Stated limits 4

What’s new here 4

Artifacts & methods 4

Topics

large language modeltool usemodel jaggednesssandboxingnetwork exfiltrationprompt engineeringbehavior analysis
← Previous AI agents are driving noisy, machine-paced attacks across cloud environments