There were plans to include measures to force social media companies to remove some "legal-but-harmful" content in the Online Safety Act, before it became law. But the proposals were scrapped over censorship concerns. Online safety campaigners argue the rules around removing harmful content needed tightening to close loopholes in the act. In January this year, Technology Secretary Peter Kyle told the BBC he had "inherited an unsatisfactory legislative settlement" in the Online Safety Act. "I'm very open-minded and I've said publicly, I think we'll have to legislate into the future again," Kyle said.

In recent years, transformer-based large language models (LLMs) have revolutionised natural language processing (NLP), with generative models opening new possibilities for tasks that require context-aware text generation. Requirements engineering (RE) has also seen a surge in the experimentation of LLMs for different tasks, including trace-link detection, regulatory compliance, and others. Requirements classification is a common task in RE. While non-generative LLMs like BERT have been successfully applied to this task, there has been limited exploration of generative LLMs. This gap raises an important question: how well can generative LLMs, which produce context-aware outputs, perform in requirements classification? In this study, we explore the effectiveness of three generative LLMs-Bloom, Gemma, and Llama-in performing both binary and multi-class requirements classification. We design an extensive experimental study involving over 400 experiments across three widely used datasets (PROMISE NFR, Functional-Quality, and SecReq). Our study concludes that while factors like prompt design and LLM architecture are universally important, others-such as dataset variations-have a more situational impact, depending on the complexity of the classification task. This insight can guide future model development and deployment strategies, focusing on optimising prompt structures and aligning model architectures with task-specific needs for improved performance.

In the field of multi-document summarization (MDS), transformer-based models have demonstrated remarkable success, yet they suffer an input length limitation. Current methods apply truncation after the retrieval process to fit the context length; however, they heavily depend on manually well-crafted queries, which are impractical to create for each document set for MDS. Additionally, these methods retrieve information at a coarse granularity, leading to the inclusion of irrelevant content. To address these issues, we propose a novel retrieval-based framework that integrates query selection and document ranking and shortening into a unified process. Our approach identifies the most salient elementary discourse units (EDUs) from input documents and utilizes them as latent queries. These queries guide the document ranking by calculating relevance scores. Instead of traditional truncation, our approach filters out irrelevant EDUs to fit the context length, ensuring that only critical information is preserved for summarization. We evaluate our framework on multiple MDS datasets, demonstrating consistent improvements in ROUGE metrics while confirming its scalability and flexibility across diverse model architectures. Additionally, we validate its effectiveness through an in-depth analysis, emphasizing its ability to dynamically select appropriate queries and accurately rank documents based on their relevance scores. These results demonstrate that our framework effectively addresses context-length constraints, establishing it as a robust and reliable solution for MDS.

Federated learning paradigm to utilize datasets across multiple data providers. In FL, cross-silo data providers often hesitate to share their high-quality dataset unless their data value can be fairly assessed. Shapley value (SV) has been advocated as the standard metric for data valuation in FL due to its desirable properties. However, the computational overhead of SV is prohibitive in practice, as it inherently requires training and evaluating an FL model across an exponential number of dataset combinations. Furthermore, existing solutions fail to achieve high accuracy and efficiency, making practical use of SV still out of reach, because they ignore choosing suitable computation scheme for approximation framework and overlook the property of utility function in FL. We first propose a unified stratified-sampling framework for two widely-used schemes. Then, we analyze and choose the more promising scheme under the FL linear regression assumption. After that, we identify a phenomenon termed key combinations, where only limited dataset combinations have a high-impact on final data value. Building on these insights, we propose a practical approximation algorithm, IPSS, which strategically selects high-impact dataset combinations rather than evaluating all possible combinations, thus substantially reducing time cost with minor approximation error. Furthermore, we conduct extensive evaluations on the FL benchmark datasets to demonstrate that our proposed algorithm outperforms a series of representative baselines in terms of efficiency and effectiveness.

Large language models (LLMs) have achieved remarkable progress, demonstrating unprecedented capabilities across various natural language processing tasks. However, the high costs associated with such exceptional performance limit the widespread adoption of LLMs, highlighting the need for prompt compression. Existing prompt compression methods primarily rely on heuristic truncation or abstractive summarization techniques, which fundamentally overlook the intrinsic mechanisms of LLMs and lack a systematic evaluation of token importance for generation. In this work, we introduce Prompt Importance Sampling (PIS), a novel compression framework that dynamically compresses prompts by sampling important tokens based on the analysis of attention scores of hidden states. PIS employs a dual-level compression mechanism: 1) at the token level, we quantify saliency using LLM-native attention scores and implement adaptive compression through a lightweight 9-layer reinforcement learning (RL) network; 2) at the semantic level, we propose a Russian roulette sampling strategy for sentence-level importance sampling. Comprehensive evaluations across multiple domain benchmarks demonstrate that our method achieves state-of-the-art compression performance. Notably, our framework serendipitously enhances reasoning efficiency through optimized context structuring. This work advances prompt engineering by offering both theoretical grounding and practical efficiency in context management for LLMs.

Streetscape Analysis with Generative AI (SAGAI): Vision-Language Assessment and Mapping of Urban Scenes Joan Perez 1 and Giovanni Fusco 2 1 Urban Geo Analytics, France 2 Universit e Cˆ ote-Azur-CNRS-AMU-Avignon Universit e, ESPACE, France April 2025 Abstract Streetscapes are an essential component of urban space. Their assessment is presently either limited to morphometric properties of their mass skeleton or requires labor-intensive qualitative evaluations of visually perceived qualities. This paper introduces SAGAI: Streetscape Analysis with Generative Artificial Intelligence, a modular workflow for scoring street-level urban scenes using open-access data and vision-language models. SAGAI integrates OpenStreetMap geometries, Google Street View imagery, and a lightweight version of the LLaVA model to generate structured spatial indicators from images via customizable natural language prompts. The pipeline includes an automated mapping module that aggregates visual scores at both the point and street levels, enabling direct cartographic interpretation. It operates without task-specific training or proprietary software dependencies, supporting scalable and interpretable analysis of urban environments. Two exploratory case studies in Nice and Vienna illustrate SAGAI's capacity to produce geospatial outputs from vision-language inference. The initial results show strong performance for binary urban-rural scene classification, moderate precision in commercial feature detection, and lower estimates, but still informative, of sidewalk width. Fully deployable by any user, SAGAI can be easily adapted to a wide range of urban research themes, such as walkability, safety, or urban design, through prompt modification alone. Keywords: Vision-Language Models, Street View Imagery, Streetscape Analysis, Geospatial AI, zero-shot inference 1 Introduction Assessing the qualities and functions of urban streetscapes is essential to understand walkability, safety, commercial vitality, and social life in cities [1, 2, 3]. However, traditional methods for observing and evaluating street-level conditions, such as field surveys, audits, and manual photo interpretation, remain time-consuming, labor-intensive, and difficult to scale beyond small pilot zones [2]. Geo-processing of vector models of the built environment allows the assessment of Email: jperez@urbangeoanalytics.com, ORCID: 0000-0003-3003-0895 Email: giovanni.fusco@univ-cotedazur.fr,

Multi-Agent Debate (MAD), leveraging collaborative interactions among Large Language Models (LLMs), aim to enhance reasoning capabilities in complex tasks. However, the security implications of their iterative dialogues and role-playing characteristics, particularly susceptibility to jailbreak attacks eliciting harmful content, remain critically underexplored. This paper systematically investigates the jailbreak vulnerabilities of four prominent MAD frameworks built upon leading commercial LLMs (GPT-4o, GPT-4, GPT-3.5-turbo, and DeepSeek) without compromising internal agents. We introduce a novel structured prompt-rewriting framework specifically designed to exploit MAD dynamics via narrative encapsulation, role-driven escalation, iterative refinement, and rhetorical obfuscation. Our extensive experiments demonstrate that MAD systems are inherently more vulnerable than single-agent setups. Crucially, our proposed attack methodology significantly amplifies this fragility, increasing average harmfulness from 28.14% to 80.34% and achieving attack success rates as high as 80% in certain scenarios. These findings reveal intrinsic vulnerabilities in MAD architectures and underscore the urgent need for robust, specialized defenses prior to real-world deployment.

Automatic extraction of definitions from legal texts is critical for enhancing the comprehension and clarity of complex legal corpora such as the United States Code (U.S.C.). We present an advanced NLP system leveraging transformer-based architectures to automatically extract defined terms, their definitions, and their scope from the U.S.C. We address the challenges of automatically identifying legal definitions, extracting defined terms, and determining their scope within this complex corpus of over 200,000 pages of federal statutory law. Building upon previous feature-based machine learning methods, our updated model employs domain-specific transformers (Legal-BERT) fine-tuned specifically for statutory texts, significantly improving extraction accuracy. Our work implements a multi-stage pipeline that combines document structure analysis with state-of-the-art language models to process legal text from the XML version of the U.S. Code. Each paragraph is first classified using a fine-tuned legal domain BERT model to determine if it contains a definition. Our system then aggregates related paragraphs into coherent definitional units and applies a combination of attention mechanisms and rule-based patterns to extract defined terms and their jurisdictional scope. The definition extraction system is evaluated on multiple titles of the U.S. Code containing thousands of definitions, demonstrating significant improvements over previous approaches. Our best model achieves 96.8% precision and 98.9% recall (98.2% F1-score), substantially outperforming traditional machine learning classifiers. This work contributes to improving accessibility and understanding of legal information while establishing a foundation for downstream legal reasoning tasks.

As machine learning increasingly influences critical domains such as credit underwriting, public policy, and talent acquisition, ensuring compliance with fairness constraints is both a legal and ethical imperative. This paper introduces a novel framework for fairness adjustments that applies to diverse machine learning tasks, including regression and classification, and accommodates a wide range of fairness metrics. Unlike traditional approaches categorized as pre-processing, in-processing, or post-processing, our method adapts in-processing techniques for use as a post-processing step. By decoupling fairness adjustments from the model training process, our framework preserves model performance on average while enabling greater flexibility in model development. Key advantages include eliminating the need for custom loss functions, enabling fairness tuning using different datasets, accommodating proprietary models as black-box systems, and providing interpretable insights into the fairness adjustments. We demonstrate the effectiveness of this approach by comparing it to Adversarial Debiasing, showing that our framework achieves a comparable fairness/accuracy tradeoff on real-world datasets.

Responsible prompt engineering has emerged as a critical framework for ensuring that generative artificial intelligence (AI) systems serve society's needs while minimizing potential harms. As generative AI applications become increasingly powerful and ubiquitous, the way we instruct and interact with them through prompts has profound implications for fairness, accountability, and transparency. This article examines how strategic prompt engineering can embed ethical and legal considerations and societal values directly into AI interactions, moving beyond mere technical optimization for functionality. This article proposes a comprehensive framework for responsible prompt engineering that encompasses five interconnected components: prompt design, system selection, system configuration, performance evaluation, and prompt management. Drawing from empirical evidence, the paper demonstrates how each component can be leveraged to promote improved societal outcomes while mitigating potential risks. The analysis reveals that effective prompt engineering requires a delicate balance between technical precision and ethical consciousness, combining the systematic rigor and focus on functionality with the nuanced understanding of social impact. Through examination of real-world and emerging practices, the article illustrates how responsible prompt engineering serves as a crucial bridge between AI development and deployment, enabling organizations to fine-tune AI outputs without modifying underlying model architectures. This approach aligns with broader "Responsibility by Design" principles, embedding ethical considerations directly into the implementation process rather than treating them as post-hoc additions. The article concludes by identifying key research directions and practical guidelines for advancing the field of responsible prompt engineering.