Explainable Artificial Intelligence (XAI) has emerged as a pillar of Trustworthy AI and aims to bring transparency in complex models that are opaque by nature. Despite the benefits of incorporating explanations in models, an urgent need is found in addressing the privacy concerns of providing this additional information to end users. In this article, we conduct a scoping review of existing literature to elicit details on the conflict between privacy and explainability. Using the standard methodology for scoping review, we extracted 57 articles from 1,943 studies published from January 2019 to December 2024. The review addresses 3 research questions to present readers with more understanding of the topic: (1) what are the privacy risks of releasing explanations in AI systems? (2) what current methods have researchers employed to achieve privacy preservation in XAI systems? (3) what constitutes a privacy preserving explanation? Based on the knowledge synthesized from the selected studies, we categorize the privacy risks and preservation methods in XAI and propose the characteristics of privacy preserving explanations to aid researchers and practitioners in understanding the requirements of XAI that is privacy compliant. Lastly, we identify the challenges in balancing privacy with other system desiderata and provide recommendations for achieving privacy preserving XAI. We expect that this review will shed light on the complex relationship of privacy and explainability, both being the fundamental principles of Trustworthy AI.

The Age-Gated Internet Is Sweeping the US. Half of the country now requires age verification to watch porn or access "harmful" content. Digital rights advocates are pushing back against legislation they say will make the internet less safe. To prove you're an adult, you may have to upload your ID or submit to an age-verifying face scan. Members of Congress considered 19 online safety bills Tuesday that may soon have a major impact on the future of the internet as age-verification laws have spread to half of the US and around the world .

Hate speech in online videos is posing an increasingly serious threat to digital platforms, especially as video content becomes increasingly multimodal and context-dependent. Existing methods often struggle to effectively fuse the complex semantic relationships between modalities and lack the ability to understand nuanced hateful content. To address these issues, we propose an innovative Reasoning-Aware Multimodal Fusion (RAMF) framework. To tackle the first challenge, we design Local-Global Context Fusion (LGCF) to capture both local salient cues and global temporal structures, and propose Semantic Cross Attention (SCA) to enable fine-grained multimodal semantic interaction. To tackle the second challenge, we introduce adversarial reasoning-a structured three-stage process where a vision-language model generates (i) objective descriptions, (ii) hate-assumed inferences, and (iii) non-hate-assumed inferences-providing complementary semantic perspectives that enrich the model's contextual understanding of nuanced hateful intent. Evaluations on two real-world hateful video datasets demonstrate that our method achieves robust generalisation performance, improving upon state-of-the-art methods by 3% and 7% in Macro-F1 and hate class recall, respectively. We will release the code after the anonymity period ends.

Auditors rely on Journal Entry Tests (JETs) to detect anomalies in tax-related ledger records, but rule-based methods generate overwhelming false positives and struggle with subtle irregularities. We investigate whether large language models (LLMs) can serve as anomaly detectors in double-entry bookkeeping. Benchmarking SoTA LLMs such as LLaMA and Gemma on both synthetic and real-world anonymized ledgers, we compare them against JETs and machine learning baselines. Our results show that LLMs consistently outperform traditional rule-based JETs and classical ML baselines, while also providing natural-language explanations that enhance interpretability. These results highlight the potential of \textbf{AI-augmented auditing}, where human auditors collaborate with foundation models to strengthen financial integrity.

As generative models become powerful, concerns around transparency, accountability, and copyright violations have intensified. Understanding how specific training data contributes to a model's output is critical. We introduce a framework for interpreting generative outputs through the automatic construction of ontologyaligned knowledge graphs (KGs). While automatic KG construction from natural text has advanced, extracting structured and ontology-consistent representations from visual content remains challenging -- due to the richness and multi-object nature of images. Leveraging multimodal large language models (LLMs), our method extracts structured triples from images, aligned with a domain-specific ontology. By comparing the KGs of generated and training images, we can trace potential influences, enabling copyright analysis, dataset transparency, and interpretable AI. We validate our method through experiments on locally trained models via unlearning, and on large-scale models through a style-specific experiment. Our framework supports the development of AI systems that foster human collaboration, creativity and stimulate curiosity.

Large language models (LLMs) are now used in settings such as Google's AI Overviews, where it summarizes multiple long documents. However, it remains unclear whether they weight all inputs equally. Focusing on abortion-related news, we construct 40 pro-neutral-con article triplets, permute each triplet into six input orders, and prompt Gemini 2.5 Flash to generate a neutral overview. We evaluate each summary against its source articles using ROUGE-L (lexical overlap), BERTScore (semantic similarity), and SummaC (factual consistency). One-way ANOVA reveals a significant primacy effect for BERTScore across all stances, indicating that summaries are more semantically aligned with the first-seen article. Pairwise comparisons further show that Position 1 differs significantly from Positions 2 and 3, while the latter two do not differ from each other, confirming a selective preference for the first document. The findings present risks for applications that rely on LLM-generated overviews and for agentic AI systems, where the steps involving LLMs can disproportionately influence downstream actions.

Abstract--The emergence of Large Language Models (LLMs) is rapidly accelerating the development of autonomous multi-agent systems (MAS), paving the way for the Internet of Agents. However, traditional centralized MAS architectures present significant challenges, including single points of failure, vulnerability to censorship, inherent scalability limitations, and critical trust issues. We propose a novel Decentralized Multi-Agent System (DMAS) architecture designed to overcome these fundamental problems by enabling trust-aware, scalable, and censorship-resistant interactions among autonomous agents. Our DMAS features a decentralized agent runtime underpinned by a blockchain-based architecture. We formalize a trust-aware communication protocol that leverages cryptographic primitives and on-chain operations to provide security properties: verifiable interaction cycles, communication integrity, authenticity, non-repudiation, and conditional confidentiality, which we further substantiate through a comprehensive security analysis. The rapid advancements in Large Language Models (LLMs) [1]-[4] have opened unprecedented avenues for creating highly autonomous and intelligent agents. These LLM-augmented agents possess remarkable capabilities in understanding natural language, performing complex reasoning, planning intricate sequences of actions, and engaging in sophisticated communication.

Generative artificial intelligence (GenAI) has garnered considerable attention for its potential utility in research and scholarship. A growing body of work in sociology and related fields demonstrates both the potential advantages and risks of GenAI, but these studies are largely proof-of-concept or specific audits of models and products. We know comparatively little about how sociologists actually use GenAI in their research practices and how they view its present and future role in the discipline. In this paper, we describe the current landscape of GenAI use in sociological research based on a survey of authors in 50 sociology journals. Our sample includes both computational sociologists and non-computational sociologists and their collaborators. We find that sociologists primarily use GenAI to assist with writing tasks: revising, summarizing, editing, and translating their own work. Respondents report that GenAI saves time and that they are curious about its capabilities, but they do not currently feel strong institutional or field-level pressure to adopt it. Overall, respondents are wary of GenAI's social and environmental impacts and express low levels of trust in its outputs, but many believe that GenAI tools will improve over the next several years. We do not find large differences between computational and non-computational scholars in terms of GenAI use, attitudes, and concern; nor do we find strong patterns by familiarity or frequency of use. We discuss what these findings suggest about the future of GenAI in sociology and highlight challenges for developing shared norms around its use in research practice.

We present Apertus, a fully open suite of large language models (LLMs) designed to address two systemic shortcomings in today's open model ecosystem: data compliance and multilingual representation. Unlike many prior models that release weights without reproducible data pipelines or regard for content-owner rights, Apertus models are pretrained exclusively on openly available data, retroactively respecting `robots.txt` exclusions and filtering for non-permissive, toxic, and personally identifiable content. To mitigate risks of memorization, we adopt the Goldfish objective during pretraining, strongly suppressing verbatim recall of data while retaining downstream task performance. The Apertus models also expand multilingual coverage, training on 15T tokens from over 1800 languages, with ~40% of pretraining data allocated to non-English content. Released at 8B and 70B scales, Apertus approaches state-of-the-art results among fully open models on multilingual benchmarks, rivalling or surpassing open-weight counterparts. Beyond model weights, we release all scientific artifacts from our development cycle with a permissive license, including data preparation scripts, checkpoints, evaluation suites, and training code, enabling transparent audit and extension.

A newly enacted New York law requires retailers to say whether your data influences the price of basic goods like a dozen eggs or toilet paper, but not how. If you're near Rochester, New York, the price for a carton of Target's Good & Gather eggs is listed as $1.99 on its website. It's unclear why the prices differ, but a new notice on Target's website offers a potential hint: "This price was set by an algorithm using your personal data." A recently enacted New York State law requires businesses that algorithmically set prices using customers' personal data to disclose that. According to the law, personal data includes any data that can be "linked or reasonably linked, directly or indirectly, with a specific consumer or device." The law doesn't require businesses to explicitly state what information about a person or device is being used or how each piece of information affects the final price a customer sees.