Researchers propose a solution to the problem of copy-pasting model-generated code into software artifacts without proper examination. They introduce 'soft barriers' that preserve access to AI assistance while making unexamined transfer less frictionless. The authors use Unicode output perturbations as an initial probe and find that output-level barriers can achieve high copy-paste resistance, but their effectiveness is highly dependent on the model and task. --- Why it matters: This research matters because it highlights a critical issue in AI-assisted programming: the ease of copying and pasting unexamined code into software artifacts. By introducing soft barriers, researchers aim to shift users from direct transfer toward editing and reconstruction, which can lead to more secure and reliable software development. Source: https://arxiv.org/abs/2608.22638