arXiv subjects

Daniele Quercia

Publications and source records attributed to Daniele Quercia.

At least 37 records · Page 2Linked to original sources

arXiv2025

The AI Model Risk Catalog: What Developers and Researchers Miss About Real-World AI Harms

We analyzed nearly 460,000 AI model cards from Hugging Face to examine how developers report risks. From these, we extracted around 3,000 unique risk mentions and built the \emph{AI Model Risk Catalog}. We compared these with risks identified by researchers in the MIT Risk Repository and with real-world incidents from the AI Incident Database. Developers focused on technical issues like bias and safety, while researchers emphasized broader social impacts. Both groups paid little attention to fraud and manipulation, which are common harms arising from how people interact with AI. Our findings show the need for clearer, structured risk reporting that helps developers think about human-interaction and systemic risks early in the design process. The catalog and paper appendix are available at: https://social-dynamics.net/ai-risks/catalog.

cs.CY↗

arXiv2025

Vitamin N: Benefits of Different Forms of Public Greenery for Urban Health

Urban greenery is often linked to better health, yet findings from past research have been inconsistent. One reason is that official greenery metrics measure the amount or nearness of greenery but ignore how often people actually may potentially see or use it in daily life. To address this gap, we introduced a new classification that separates on-road greenery, which people see while walking through streets, from off-road greenery, which requires planned visits. We did so by combining aerial imagery of Greater London and greenery data from OpenStreetMap with quantified greenery from over 100,000 Google Street View images and accessibility estimates based on 160,000 road segments. We linked these measures to 7.45 billion medical prescriptions issued by the National Health Service and processed through our methodology. These prescriptions cover five conditions: diabetes, hypertension, asthma, depression, and anxiety, as well as opioid use. As hypothesized, we found that green on-road was more strongly linked to better health than four widely used official measures. For example, hypertension prescriptions dropped by 3.68% in wards with on-road greenery above the median citywide level compared to those below it. If all below-median wards reached the citywide median in on-road greenery, prescription costs could fall by up to £3.15 million each year. These results suggest that greenery seen in daily life may be more relevant than public yet secluded greenery, and that official metrics commonly used in the literature have important limitations.

cs.CY↗

arXiv2025

The Potential Impact of Disruptive AI Innovations on U.S. Occupations

The rapid rise of AI is poised to disrupt the labor market. However, AI is not a monolith; its impact depends on both the nature of the innovation and the jobs it affects. While computational approaches are emerging, there is no consensus on how to systematically measure an innovation's disruptive potential. Here, we calculate the disruption index of 3,237 U.S. AI patents (2015-2022) and link them to job tasks to distinguish between "consolidating" AI innovations that reinforce existing structures and "disruptive" AI innovations that alter them. Our analysis reveals that consolidating AI primarily targets physical, routine, and solo tasks, common in manufacturing and construction in the Midwest and central states. By contrast, disruptive AI affects unpredictable and mental tasks, particularly in coastal science and technology sectors. Surprisingly, we also find that disruptive AI disproportionately affects areas already facing skilled labor shortages, suggesting disruptive AI technologies may accelerate change where workers are scarce rather than replacing a surplus. Ultimately, consolidating AI appears to extend current automation trends, while disruptive AI is set to transform complex mental work, with a notable exception for collaborative tasks.

cs.CY↗

arXiv2025

The Experience of Running: Recommending Routes Using Sensory Mapping in Urban Environments

Depending on the route, runners may experience frustration, freedom, or fulfilment. However, finding routes that are conducive to the psychological experience of running remains an unresolved task in the literature. In a mixed-method study, we interviewed 7 runners to identify themes contributing to running experience, and quantitatively examined these themes in an online survey with 387 runners. Using Principal Component Analysis on the survey responses, we developed a short experience sampling questionnaire that captures the three most important dimensions of running experience: \emph{performance \& achievement}, \emph{environment}, and \emph{mind \& social connectedness}. Using path preferences obtained from the online survey, we clustered them into two types of routes: \emph{scenic} (associated with nature and greenery) and \emph{urban} (characterized by the presence of people); and developed a routing engine for path recommendations. We discuss challenges faced in developing the routing engine, and provide guidelines to integrate it into mobile and wearable running apps.

cs.HC↗

arXiv2025

AI, Jobs, and the Automation Trap: Where Is HCI?

As artificial intelligence (AI) continues to reshape the workforce, its current trajectory raises pressing questions about its ultimate purpose. Why does job automation dominate the agenda, even at the expense of human agency and equity? This paper critiques the automation-centric paradigm, arguing that current reward structures, which largely focus on cost reduction, drive the overwhelming emphasis on task replacement in AI patents. Meanwhile, Human-Centered AI (HCAI), which envisions AI as a collaborator augmenting human capabilities and aligning with societal values, remains a fugitive from the mainstream narrative. Despite its promise, HCAI has gone ``missing'', with little evidence of its principles translating into patents or real-world impact. To increase impact, actionable interventions are needed to disrupt existing incentive structures within the HCI community. We call for a shift in priorities to support translational research, foster cross-disciplinary collaboration, and promote metrics that reward tangible and real-world impact.

cs.HC↗

arXiv2025

RiskRAG: A Data-Driven Solution for Improved AI Model Risk Reporting

Risk reporting is essential for documenting AI models, yet only 14% of model cards mention risks, out of which 96% copying content from a small set of cards, leading to a lack of actionable insights. Existing proposals for improving model cards do not resolve these issues. To address this, we introduce RiskRAG, a Retrieval Augmented Generation based risk reporting solution guided by five design requirements we identified from literature, and co-design with 16 developers: identifying diverse model-specific risks, clearly presenting and prioritizing them, contextualizing for real-world uses, and offering actionable mitigation strategies. Drawing from 450K model cards and 600 real-world incidents, RiskRAG pre-populates contextualized risk reports. A preliminary study with 50 developers showed that they preferred RiskRAG over standard model cards, as it better met all the design requirements. A final study with 38 developers, 40 designers, and 37 media professionals showed that RiskRAG improved their way of selecting the AI model for a specific application, encouraging a more careful and deliberative decision-making. The RiskRAG project page is accessible at: https://social-dynamics.net/ai-risks/card.

cs.SE↗

arXiv2025

The Hall of AI Fears and Hopes: Comparing the Views of AI Influencers and those of Members of the U.S. Public Through an Interactive Platform

AI development is shaped by academics and industry leaders - let us call them ``influencers'' - but it is unclear how their views align with those of the public. To address this gap, we developed an interactive platform that served as a data collection tool for exploring public views on AI, including their fears, hopes, and overall sense of hopefulness. We made the platform available to 330 participants representative of the U.S. population in terms of age, sex, ethnicity, and political leaning, and compared their views with those of 100 AI influencers identified by Time magazine. The public fears AI getting out of control, while influencers emphasize regulation, seemingly to deflect attention from their alleged focus on monetizing AI's potential. Interestingly, the views of AI influencers from underrepresented groups such as women and people of color often differ from the views of underrepresented groups in the public.

cs.HC↗

arXiv2025

C3AI: Crafting and Evaluating Constitutions for Constitutional AI

Constitutional AI (CAI) guides LLM behavior using constitutions, but identifying which principles are most effective for model alignment remains an open challenge. We introduce the C3AI framework (\textit{Crafting Constitutions for CAI models}), which serves two key functions: (1) selecting and structuring principles to form effective constitutions before fine-tuning; and (2) evaluating whether fine-tuned CAI models follow these principles in practice. By analyzing principles from AI and psychology, we found that positively framed, behavior-based principles align more closely with human preferences than negatively framed or trait-based principles. In a safety alignment use case, we applied a graph-based principle selection method to refine an existing CAI constitution, improving safety measures while maintaining strong general reasoning capabilities. Interestingly, fine-tuned CAI models performed well on negatively framed principles but struggled with positively framed ones, in contrast to our human alignment results. This highlights a potential gap between principle design and model adherence. Overall, C3AI provides a structured and scalable approach to both crafting and evaluating CAI constitutions.

cs.AI↗

arXiv2025

Atlas of AI Risks: Enhancing Public Understanding of AI Risks

The prevailing methodologies for visualizing AI risks have focused on technical issues such as data biases and model inaccuracies, often overlooking broader societal risks like job loss and surveillance. Moreover, these visualizations are typically designed for tech-savvy individuals, neglecting those with limited technical skills. To address these challenges, we propose the Atlas of AI Risks-a narrative-style tool designed to map the broad risks associated with various AI technologies in a way that is understandable to non-technical individuals as well. To both develop and evaluate this tool, we conducted two crowdsourcing studies. The first, involving 40 participants, identified the design requirements for visualizing AI risks for decision-making and guided the development of the Atlas. The second study, with 140 participants reflecting the US population in terms of age, sex, and ethnicity, assessed the usability and aesthetics of the Atlas to ensure it met those requirements. Using facial recognition technology as a case study, we found that the Atlas is more user-friendly than a baseline visualization, with a more classic and expressive aesthetic, and is more effective in presenting a balanced assessment of the risks and benefits of facial recognition. Finally, we discuss how our design choices make the Atlas adaptable for broader use, allowing it to generalize across the diverse range of technology applications represented in a database that reports various AI incidents.

cs.HC↗

arXiv2024

The Impact of Responsible AI Research on Innovation and Development

Translational research, especially in the fast-evolving field of Artificial Intelligence (AI), is key to converting scientific findings into practical innovations. In Responsible AI (RAI) research, translational impact is often viewed through various pathways, including research papers, blogs, news articles, and the drafting of forthcoming AI legislation (e.g., the EU AI Act). However, the real-world impact of RAI research remains an underexplored area. Our study aims to capture it through two pathways: \emph{patents} and \emph{code repositories}, both of which provide a rich and structured source of data. Using a dataset of 200,000 papers from 1980 to 2022 in AI and related fields, including Computer Vision, Natural Language Processing, and Human-Computer Interaction, we developed a Sentence-Transformers Deep Learning framework to identify RAI papers. This framework calculates the semantic similarity between paper abstracts and a set of RAI keywords, which are derived from the NIST's AI Risk Management Framework; a framework that aims to enhance trustworthiness considerations in the design, development, use, and evaluation of AI products, services, and systems. We identified 1,747 RAI papers published in top venues such as CHI, CSCW, NeurIPS, FAccT, and AIES between 2015 and 2022. By analyzing these papers, we found that a small subset that goes into patents or repositories is highly cited, with the translational process taking between 1 year for repositories and up to 8 years for patents. Interestingly, impactful RAI research is not limited to top U.S. institutions, but significant contributions come from European and Asian institutions. Finally, the multidisciplinary nature of RAI papers, often incorporating knowledge from diverse fields of expertise, was evident as these papers tend to build on unconventional combinations of prior knowledge.

cs.HC↗

arXiv2024

Co-designing an AI Impact Assessment Report Template with AI Practitioners and AI Compliance Experts

In the evolving landscape of AI regulation, it is crucial for companies to conduct impact assessments and document their compliance through comprehensive reports. However, current reports lack grounding in regulations and often focus on specific aspects like privacy in relation to AI systems, without addressing the real-world uses of these systems. Moreover, there is no systematic effort to design and evaluate these reports with both AI practitioners and AI compliance experts. To address this gap, we conducted an iterative co-design process with 14 AI practitioners and 6 AI compliance experts and proposed a template for impact assessment reports grounded in the EU AI Act, NIST's AI Risk Management Framework, and ISO 42001 AI Management System. We evaluated the template by producing an impact assessment report for an AI-based meeting companion at a major tech company. A user study with 8 AI practitioners from the same company and 5 AI compliance experts from industry and academia revealed that our template effectively provides necessary information for impact assessments and documents the broad impacts of AI systems. Participants envisioned using the template not only at the pre-deployment stage for compliance but also as a tool to guide the design stage of AI uses.

cs.HC↗

arXiv2024

The Potential Impact of AI Innovations on U.S. Occupations

An occupation is comprised of interconnected tasks, and it is these tasks, not occupations themselves, that are affected by AI. To evaluate how tasks may be impacted, previous approaches utilized manual annotations or coarse-grained matching. Leveraging recent advancements in machine learning, we replace coarse-grained matching with more precise deep learning approaches. Introducing the AI Impact (AII) measure, we employ Deep Learning Natural Language Processing to automatically identify AI patents that may impact various occupational tasks at scale. Our methodology relies on a comprehensive dataset of 17,879 task descriptions and quantifies AI's potential impact through analysis of 24,758 AI patents filed with the United States Patent and Trademark Office (USPTO) between 2015 and 2022. Our results reveal that some occupations will potentially be impacted, and that impact is intricately linked to specific skills. These include not only routine tasks (codified as a series of steps), as previously thought, but also non-routine ones (e.g., diagnosing health conditions, programming computers, and tracking flight routes). However, AI's impact on labour is limited by the fact that some of the occupations affected are augmented rather than replaced (e.g., neurologists, software engineers, air traffic controllers), and the sectors affected are experiencing labour shortages (e.g., IT, Healthcare, Transport).

cs.CY↗

arXiv2024

Good Intentions, Risky Inventions: A Method for Assessing the Risks and Benefits of AI in Mobile and Wearable Uses

Integrating Artificial Intelligence (AI) into mobile and wearables offers numerous benefits at individual, societal, and environmental levels. Yet, it also spotlights concerns over emerging risks. Traditional assessments of risks and benefits have been sporadic, and often require costly expert analysis. We developed a semi-automatic method that leverages Large Language Models (LLMs) to identify AI uses in mobile and wearables, classify their risks based on the EU AI Act, and determine their benefits that align with globally recognized long-term sustainable development goals; a manual validation of our method by two experts in mobile and wearable technologies, a legal and compliance expert, and a cohort of nine individuals with legal backgrounds who were recruited from Prolific, confirmed its accuracy to be over 85\%. We uncovered that specific applications of mobile computing hold significant potential in improving well-being, safety, and social equality. However, these promising uses are linked to risks involving sensitive data, vulnerable groups, and automated decision-making. To avoid rejecting these risky yet impactful mobile and wearable uses, we propose a risk assessment checklist for the Mobile HCI community.

cs.HC↗

arXiv2024

The Atlas of AI Incidents in Mobile Computing: Visualizing the Risks and Benefits of AI Gone Mobile

Today's visualization tools for conveying the risks and benefits of AI technologies are largely tailored for those with technical expertise. To bridge this gap, we have developed a visualization that employs narrative patterns and interactive elements, enabling the broader public to gradually grasp the diverse risks and benefits associated with AI. Using a dataset of 54 real-world incidents involving AI in mobile computing, we examined design choices that enhance public understanding and provoke reflection on how certain AI applications - even those deemed low-risk by law - can still lead to significant incidents. Visualization: https://social-dynamics.net/mobile-ai-risks

cs.HC↗

arXiv2024

ExploreGen: Large Language Models for Envisioning the Uses and Risks of AI Technologies

Responsible AI design is increasingly seen as an imperative by both AI developers and AI compliance experts. One of the key tasks is envisioning AI technology uses and risks. Recent studies on the model and data cards reveal that AI practitioners struggle with this task due to its inherently challenging nature. Here, we demonstrate that leveraging a Large Language Model (LLM) can support AI practitioners in this task by enabling reflexivity, brainstorming, and deliberation, especially in the early design stages of the AI development process. We developed an LLM framework, ExploreGen, which generates realistic and varied uses of AI technology, including those overlooked by research, and classifies their risk level based on the EU AI Act regulation. We evaluated our framework using the case of Facial Recognition and Analysis technology in nine user studies with 25 AI practitioners. Our findings show that ExploreGen is helpful to both developers and compliance experts. They rated the uses as realistic and their risk classification as accurate (94.5%). Moreover, while unfamiliar with many of the uses, they rated them as having high adoption potential and transformational impact.

cs.HC↗

arXiv2024

NLPGuard: A Framework for Mitigating the Use of Protected Attributes by NLP Classifiers

AI regulations are expected to prohibit machine learning models from using sensitive attributes during training. However, the latest Natural Language Processing (NLP) classifiers, which rely on deep learning, operate as black-box systems, complicating the detection and remediation of such misuse. Traditional bias mitigation methods in NLP aim for comparable performance across different groups based on attributes like gender or race but fail to address the underlying issue of reliance on protected attributes. To partly fix that, we introduce NLPGuard, a framework for mitigating the reliance on protected attributes in NLP classifiers. NLPGuard takes an unlabeled dataset, an existing NLP classifier, and its training data as input, producing a modified training dataset that significantly reduces dependence on protected attributes without compromising accuracy. NLPGuard is applied to three classification tasks: identifying toxic language, sentiment analysis, and occupation classification. Our evaluation shows that current NLP classifiers heavily depend on protected attributes, with up to $23\%$ of the most predictive words associated with these attributes. However, NLPGuard effectively reduces this reliance by up to $79\%$, while slightly improving accuracy.

cs.CL↗

arXiv2024

WEIRD ICWSM: How Western, Educated, Industrialized, Rich, and Democratic is Social Computing Research?

Much of the research in social computing analyzes data from social media platforms, which may inherently carry biases. An overlooked source of such bias is the over-representation of WEIRD (Western, Educated, Industrialized, Rich, and Democratic) populations, which might not accurately mirror the global demographic diversity. We evaluated the dependence on WEIRD populations in research presented at the AAAI ICWSM conference; the only venue whose proceedings are fully dedicated to social computing research. We did so by analyzing 494 papers published from 2018 to 2022, which included full research papers, dataset papers and posters. After filtering out papers that analyze synthetic datasets or those lacking clear country of origin, we were left with 420 papers from which 188 participants in a crowdsourcing study with full manual validation extracted data for the WEIRD scores computation. This data was then used to adapt existing WEIRD metrics to be applicable for social media data. We found that 37% of these papers focused solely on data from Western countries. This percentage is significantly less than the percentages observed in research from CHI (76%) and FAccT (84%) conferences, suggesting a greater diversity of dataset origins within ICWSM. However, the studies at ICWSM still predominantly examine populations from countries that are more Educated, Industrialized, and Rich in comparison to those in FAccT, with a special note on the 'Democratic' variable reflecting political freedoms and rights. This points out the utility of social media data in shedding light on findings from countries with restricted political freedoms. Based on these insights, we recommend extensions of current "paper checklists" to include considerations about the WEIRD bias and call for the community to broaden research inclusivity by encouraging the use of diverse datasets from underrepresented regions.

cs.HC↗

arXiv2024

RAI Guidelines: Method for Generating Responsible AI Guidelines Grounded in Regulations and Usable by (Non-)Technical Roles

Many guidelines for responsible AI have been suggested to help AI practitioners in the development of ethical and responsible AI systems. However, these guidelines are often neither grounded in regulation nor usable by different roles, from developers to decision makers. To bridge this gap, we developed a four-step method to generate a list of responsible AI guidelines; these steps are: (1) manual coding of 17 papers on responsible AI; (2) compiling an initial catalog of responsible AI guidelines; (3) refining the catalog through interviews and expert panels; and (4) finalizing the catalog. To evaluate the resulting 22 guidelines, we incorporated them into an interactive tool and assessed them in a user study with 14 AI researchers, engineers, designers, and managers from a large technology company. Through interviews with these practitioners, we found that the guidelines were grounded in current regulations and usable across roles, encouraging self-reflection on ethical considerations at early stages of development. This significantly contributes to the concept of `Responsible AI by Design' -- a design-first approach that embeds responsible AI values throughout the development lifecycle and across various business roles.

cs.HC↗