Social network analysis is known to provide a wealth of insights relevant to many aspects of policymaking. Yet, the social data needed to construct social networks are not always available. Furthermore, even when they are, interpreting such networks often relies on extraneous knowledge. Here, we propose an approach to infer social networks directly from the texts produced by actors and the terminological similarities that these texts exhibit. This approach relies on fitting a topic model to the texts produced by these actors and measuring topic profile correlations between actors. This reveals what can be called “hidden communities of interest,” that is, groups of actors sharing similar semantic contents but whose social relationships with one another may be unknown or underlying. Network interpretation follows from the topic model. Diachronic perspectives can also be built by modeling the networks over different time periods and mapping genealogical relationships between communities. As a case study, the approach is deployed over a working corpus of academic articles (domain of philosophy of science; N=16,917).

This paper explores citizens’ stances toward the use of artificial intelligence (AI) in public services in Norway. Utilizing a social contract perspective, the study analyzes the government–citizen relationship at macro, meso, and micro levels. A prototype of an AI-enabled public welfare service was designed and presented to 20 participants who were interviewed to investigate their stances on the described AI use. We found a generally positive attitude and identified three factors contributing to this: (a) the high level of trust in government (macro level); (b) the balanced value proposition between individual and collective needs (meso level); and (c) the reassurance provided by having humans in the loop and providing transparency into processes, data, and model’s logic (microlevel). The findings provide valuable insights into citizens’ stances for socially responsible AI in public services. These insights can inform policy and guide the design and implementation of AI systems in the public sector by foregrounding the government–citizen relationship.

We apply moral foundations theory (MFT) to explore how the public conceptualizes the first eight months of the conflict between Ukraine and the Russian Federation (Russia). Our analysis includes over 1.1 million English tweets related to the conflict over the first 36 weeks. We used linguistic inquiry word count (LIWC) and a moral foundations dictionary to identify tweets’ moral components (care, fairness, loyalty, authority, and sanctity) from the United States, pre- and post-Cold War NATO countries, Ukraine, and Russia. Following an initial spike at the beginning of the conflict, tweet volume declined and stabilized by week 10. The level of moral content varied significantly across the five regions and the five moral components. Tweets from the different regions included significantly different moral foundations to conceptualize the conflict. Across all regions, tweets were dominated by loyalty content, while fairness content was infrequent. Moral content over time was relatively stable, and variations were linked to reported conflict events.

Intergovernmental collaboration is needed to address global problems. Modern solutions to these problems often include data-driven methods like artificial intelligence (AI), which require large amounts of data to perform well. As AI emerges as a central catalyst in deriving effective solutions for global problems, the infrastructure that supports its data needs becomes crucial. However, data sharing between governments is often constrained due to socio-technical barriers such as concerns over data privacy, data sovereignty issues, and the risks of information misuse. Federated learning (FL) presents a promising solution as a decentralized AI methodology, enabling the use of data from multiple silos without necessitating central aggregation. Instead of sharing raw data, governments can build their own models and just share the model parameters with a central server aggregating all parameters, resulting in a superior overall model. By conducting a structured literature review, we show how major intergovernmental data-sharing challenges listed by the Organisation for Economic Co-operation and Development can be overcome by utilizing FL. Furthermore, we provide a tangible resource implementing FL linked to the Ukrainian refugee crisis that can be utilized by researchers and policymakers alike who want to implement FL in cases where data cannot be shared. Enhanced AI while maintaining privacy through FL thus allows governments to collaboratively address global problems, positively impacting governments and citizens.

The capabilities of large language models (LLMs) have advanced to the point where entire textbooks can be queried using retrieval-augmented generation (RAG), enabling AI to integrate external, up-to-date information into its responses. This study evaluates the ability of two OpenAI models, GPT-3.5 Turbo and GPT-4 Turbo, to create and answer exam questions based on an undergraduate textbook. 14 exams were created with four true-false, four multiple-choice, and two short-answer questions derived from an open-source Pacific Studies textbook. Model performance was evaluated with and without access to the source material using text-similarity metrics such as ROUGE-1, cosine similarity, and word embeddings. Fifty-six exam scores were analyzed, revealing that RAG-assisted models significantly outperformed those relying solely on pre-trained knowledge. GPT-4 Turbo also consistently outperformed GPT-3.5 Turbo in accuracy and coherence, especially in short-answer responses. These findings demonstrate the potential of LLMs in automating exam generation while maintaining assessment quality. However, they also underscore the need for policy frameworks that promote fairness, transparency, and accessibility. Given regulatory considerations outlined in the European Union AI Act and the NIST AI Risk Management Framework, institutions using AI in education must establish governance protocols, bias mitigation strategies, and human oversight measures. The results of this study contribute to ongoing discussions on responsibly integrating AI in education, advocating for institutional policies that support AI-assisted assessment while preserving academic integrity. The empirical results suggest not only performance benefits but also actionable governance mechanisms, such as verifiable retrieval pipelines and oversight protocols, that can guide institutional policies.

The rise of visually driven platforms like Instagram has reshaped how information is shared and understood. This study examines the role of social, cultural, and political (SCP) symbols in Instagram posts during Taiwan’s 2024 election, focusing on their influence in anti-misinformation efforts. Using large language models (LLMs)—GPT-4 Omni and Gemini Pro Vision—we analyzed thousands of posts to extract and classify symbolic elements, comparing model performance in consistency and interpretive depth. We evaluated how SCP symbols affect user engagement, perceptions of fairness, and content spread. Engagement was measured by likes, while diffusion patterns followed the SEIZ epidemiological model. Findings show that posts featuring SCP symbols consistently received more interaction, even when follower counts were equal. Although political content creators often had larger audiences, posts with cultural symbols drove the highest engagement, were perceived as more fair and trustworthy, and spread more rapidly across networks. Our results suggest that symbolic richness influences online interactions more than audience size. By integrating semiotic analysis, LLM-based interpretation, and diffusion modeling, this study offers a novel framework for understanding how symbolic communication shapes engagement on visual platforms. These insights can guide designers, policymakers, and strategists in developing culturally resonant, symbol-aware messaging to combat misinformation and promote credible narratives.

Over 193 countries have signed at least one of more than 500 multilateral treaties addressing critical global issues, such as human rights, environmental protection, and trade. Ratifying a treaty obligates a country, as a “State Party,” to report to the United Nations on its progress toward implementing the treaty’s provisions. These reports and their associated review processes generate a wealth of textual data. Effectively monitoring, reviewing, and assessing national, regional, and global progress toward these treaty commitments is crucial for ensuring compliance and realizing the benefits of international cooperation. The UN Convention on the Rights of Persons with Disabilities (CRPD), which has been ratified by 191 countries, exemplifies this challenge. With over 1.3 billion people worldwide living with disabilities, the CRPD aims to promote a shift from a charity-based “medical model” that views disability as an individual deficiency, to a rights-based “social justice model” that emphasizes societal barriers and inclusivity. Each State Party submits periodic reports to the Committee on the Rights of Persons with Disabilities detailing their implementation efforts. This study analyzed all available CRPD State Reports (N = 170) using text mining, Natural Language Processing, and GenerativeAI tools to assess global progress, identify regional variations, and explore the factors influencing successful implementation. The findings reveal evidence of widespread CRPD implementation, growing support for social justice and economic inclusion, and the importance of civil society engagement. Hybrid data analysis approach of this study offers a promising framework for harnessing the power of textual data to advance the realization of treaty commitments worldwide.

As social media continues to grow, understanding the impact of storytelling on stakeholder engagement becomes increasingly important for policymakers and organizations who wish to influence policymaking. While prior research has explored narrative strategies in advertising and branding, researchers have paid scant attention to the specific influence of stories on social media stakeholder engagement. This study addresses this gap by employing Narrative Transportation Theory (NTT) and leveraging Natural Language Processing (NLP) to analyze the intricate textual data generated by social media platforms. The analysis of 85,075 Facebook publications from leading Canadian manufacturing companies, using Spearman’s rank correlation coefficient, underscores that individual storytelling components—character, sequence of events, and setting—along with the composite narrative structure significantly enhance stakeholder engagement. This research contributes to a deeper understanding of storytelling dynamics in social media, emphasizing the importance of crafting compelling stories to drive meaningful stakeholder engagement in the digital realm. The results of our research can prove useful for those who wish to influence policymakers or for policymakers who want to promote new policies.

The emergence of large language models has significantly expanded the use of natural language processing (NLP), even as it has heightened exposure to adversarial threats. We present an overview of adversarial NLP with an emphasis on challenges, policy implications, emerging areas, and future directions. First, we review attack methods and evaluate the vulnerabilities of popular NLP models. Then, we review defense strategies that include adversarial training. We describe major policy implications, identify key trends, and suggest future directions, such as the use of Bayesian methods to improve the security and robustness of NLP systems.

This study investigates unintended information flow in large language models (LLMs) by proposing a computational linguistic framework for detecting and analyzing domain anchorage. Domain anchorage is a phenomenon potentially caused by in-context learning or latent “cache” retention of prior inputs, which enables language models to infer and reinforce shared latent concepts across interactions, leading to uniformity in responses that can persist across distinct users or prompts. Using GPT-4 as a case study, our framework systematically quantifies the lexical, syntactic, semantic, and positional similarities between inputs and outputs to detect these domain anchorage effects. We introduce a structured methodology to evaluate the associated risks and highlight the need for robust mitigation strategies. By leveraging domain-aware analysis, this work provides a scalable framework for monitoring information persistence in LLMs, which can inform enterprise guardrails to ensure response consistency, privacy, and safety in real-world deployments.

National digital ID apps are increasingly gaining popularity globally. As how we transact in the world is increasingly mediated by the digital, questions need to be asked about how these apps support the inclusion of disabled people. In particular, international instruments, such as the United Nations Convention on the Rights of Persons with Disabilities, spotlight the need for inclusive information and communication technologies. In this paper, we adopt a critical disability studies lens to analyse the workings of state-designed digital IDs—Singpass app—and what they can tell us about existing ways of designing for digital inclusion. We situate the case of the Singpass app within the rise of global digital transactions and the political-technical infrastructures that shape their accessibility. We analyse the ways Singpass centres disability, the problems it may still entail, and the possible implications for inclusion. At the same time, we uncover the lessons Singpass’s development holds for questions of global digital inclusion.

In our digitalized modern society where cyber-physical systems and internet-of-things (IoT) devices are increasingly commonplace, it is paramount that we are able to assure the cybersecurity of the systems that we rely on. As a fundamental policy, we join the advocates of multilayered cybersecurity measures, where resilience is built into IoT systems by relying on multiple defensive techniques. While existing legislation such as the General Data Protection Regulation (GDPR) also takes this stance, the technical implementation of these measures is left open. This invites research into the landscape of multilayered defensive measures, and within this problem space, we focus on two defensive measures: obfuscation and diversification. In this study, through a literature review, we situate these measures within the broader IoT cybersecurity landscape and show how they operate with other security measures built on the network and within IoT devices themselves. Our findings highlight that obfuscation and diversification show promise in contributing to a cost-effective robust cybersecurity ecosystem in today’s diverse cyber threat landscape.