Output-level Unicode soft barriers can achieve high copy-paste resistance, but effectiveness is highly model- and task-dependent¶

Claim · All claims
Evidence · 1 study · causal r1 · q2
Currency · no dated study recorded · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Across HumanEval and MBPP with four LLMs and four perturbation families, CPR varies from near zero to near perfect, so no universal Unicode barrier exists. → cpr-results

Evidence¶

cpr-results¶

Iyiola E. Olatunji, Alberick Euraste Djire, Jacques Klein, and Tegawendé F. Bissyandé. 2026. Do Not Copy/Paste: Soft Barriers for Copying in AI-Assisted Programming. Proceedings of the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE '26). https://doi.org/10.1145/3832783.3834554

q2 · i? · causal · r1

Automated evaluation of four perturbation families across Claude Sonnet 4.5, DeepSeek-V3, GPT-5.2, and GPT-5.2 Codex on HumanEval and MBPP, reported in Table 3. The paper states "CPR varies widely, from near zero to near perfect" and concludes a combination of barriers would be more effective for deployment.

"CPR varies widely, from near zero to near perfect, showing that soft-barrier effectiveness is highly dependent on the model and perturbation strategy."

Discussion¶

Related Claims¶

• In an exploratory pilot, a soft barrier shifted participants from direct copying toward editing, understanding, and iterative resubmission without preventing task completion — a narrower finding that bears on this claim
• Pass@1 is not a sufficient success metric for soft barriers because it conflates successful barriers with model generation failures — related
• Automatically updating the flowchart and injecting generated code into editors drew participants away from explanations and toward direct code editing — related