Legal documents are often verbose and structurally complex, posing significant barriers to public understanding and equitable access to justice. Despite growing interest in text simplification, efforts targeting the legal domain remain limited by a lack of robust, high-quality resources. In this paper, we address this gap by introducing SIMPLE-LAW, a curated benchmark dataset of over 6,000 aligned pairs of original and simplified legal passages, specifically constructed to facilitate research in legal text simplification by leveraging large language models (LLMs). We evaluate this dataset across both in-context learning and parameter-efficient fine-tuning paradigms using a range of state-of-the-art LLMs, with Unsloth variants of Mistral, LLaMA-3.2, Gemma, and Qwen-2.5. We assess performance using BERTScore, ROUGE, SARI, and a hallucination detection score, to capture both fidelity and readability. Results show that fine-tuned models significantly outperform in-context learners in terms of simplification quality and factual consistency. By offering a new dataset, rigorous evaluation, and baseline comparisons, our work provides a critical foundation for developing transparent and accessible AI systems in the legal domain.

Human annotation is the empirical foundation of much NLP research, from dataset construction to model evaluation, but papers often leave unclear who produced the annotations and how the annotation process was controlled. We provide the first large-scale, task-level audit of human annotation reporting across major NLP venues, asking which annotation details are documented, which are missing, and how reporting varies across time, topic, venue, and intended use of human judgment. We introduce a unified taxonomy of annotation-reporting practices and validate an LLM-assisted extraction pipeline against Annotated_gold, a human-adjudicated gold standard of 41 papers and 72 annotation tasks, where the best model reaches human-comparable agreement with adjudicated labels, with Krippendorff’s alpha = 0.606 versus 0.585 for human–human agreement. Using this pipeline, we construct Annotated_llm, a dataset covering ACL-venue papers from 2018–2025, with 2,667 extracted annotation tasks from 1,603 papers, and find that papers frequently report operational details such as recruitment strategies, annotator expertise, and annotation volume, but often omit details needed to assess annotation validity, including training, language proficiency, compensation, socio-demographics, adjudication, and agreement values, especially in model-evaluation studies. Our results show that annotation reporting in NLP has improved over time but remains uneven, and they establish a scalable framework and bare-minimum reporting recommendations for making human annotation more reliable, reproducible, and interpretable.