67471 research outputs found
Sort by
Agentic AI for Financial Crime Compliance
Get PDFThe cost and complexity of financial crime compliance (FCC) continue to rise, often without measurable improvements in effectiveness. While AI offers potential, most solutions remain opaque and poorly aligned with regulatory expectations. This paper presents the design and deployment of an agentic AI system for FCC in digitally native financial platforms. Developed through an Action Design Research (ADR) process with a fintech firm and regulatory stakeholders, the system automates onboarding, monitoring, investigation, and reporting, emphasizing explainability, traceability, and compliance-by-design. Using artifact-centric modeling, it assigns clearly bounded roles to autonomous agents and enables task-specific model routing and audit logging. The contribution includes a reference architecture, a real-world prototype, and insights into how Agentic AI can reconfigure FCC workflows under regulatory constraints. Our findings extend IS literature on AI-enabled compliance by demonstrating how automation, when embedded within accountable governance structures, can support transparency and institutional trust in high-stakes, regulated environments
Privacy Concerns about AI Smart Glasses: A Cross-Cultural Study of the U.S. and South Korea
Get PDFNewer AI smart glasses (AISG) enable users to unobtrusively capture photographs and videos, raising privacy issues for individuals who may remain unaware that they are being recorded. This cross-cultural online survey examined privacy perspectives between two technologically sophisticated yet culturally distinct societies, the United States and South Korea. The study investigated how bystander privacy considerations could guide AI smart glasses technology development and shape public opinions regarding technology governance. As predicted, South Koreans reported significantly greater privacy concerns, agreement with social norms supporting bystanders’ privacy rights, support for self-regulation, and support for both legal restrictions and technological solutions concerning the use of AISG. In both countries, individuals’ own privacy concerns were also positively correlated with these factors. Women in both countries exhibited greater privacy concerns. Participants’ age increased privacy concerns in the U.S., but age decreased concerns in South Korea
Investigating the Impact of Rewards and Sanctions on Developers’ Proactive AI Accountability Behavior
Get PDFAccountability of artificial intelligence (AI)-based systems is often addressed reactively, mainly after harm occurs. This study shifts toward proactive approaches, highlighting AI developers’ role in risk mitigation. Proactive AI accountability behavior refers to self-initiated, future-oriented actions that go beyond formal job roles to justify developers’ actions and decisions, and to facilitate the clear attribution of accountability. Drawing on Proactive Motivation Theory, we conducted an online experiment (n = 264) to investigate how governance mechanisms (rewards vs. sanctions) and motivational states impact such behavior. Our results reveal flexible role orientation as the key driver of proactive behavior and how rewards and sanctions impact such a mindset. We contribute by conceptualizing proactive AI accountability behavior and providing a theoretical model that explains its emergence, underscoring the importance of using rewards to foster a proactive mindset alongside sanctions as guardrails against harmful initiatives
Introduction to the Minitrack on Tools and Processes for Enabling Agile Projects, Teams, and Organizations
No full text
A Scalable Distillation and Metric-Learning Pipeline for Adaptive Weed Classification on Edge Devices
Get PDFSmart agriculture increasingly relies on automated weed detection to reduce inputs and labor. Deploying deep learning on edge devices is difficult due to limited compute and evolving weed classes. We propose a pipeline that combines partial fine-tuning of an EfficientNet-B7 teacher, embedding level distillation into lightweight students (MobileNetV3, ShuffleNetV2, EfficientNet-B0), and semi-hard triplet metric learning. The system learns 2048D embeddings and is evaluated with N-way/K-shot episodes to mimic a few-label condition. Dynamic INT8 quantization enables CPU-only inference with minimal accuracy loss. The approach adapts rapidly to novel species with few labels while meeting real-time edge constraints, supporting sustainable herbicide management in practice
Self-Training Approach for Smishing Detection in Multilingual and Low-Resource Settings
Get PDFSmishing--phishing conducted via SMS--continues to spread globally. Most detection systems are built using only English data, which limits their use for other languages. We propose a multilingual smishing detection framework based on self-training with XLM-RoBERTa. Starting with English-labeled data, we apply zero-shot inference to Bengali and Swahili messages, extract high-confidence predictions, and incorporate them as pseudo-labeled examples for fine-tuning. Our approach improves recall and F1 scores in both target languages without requiring parallel corpora or costly annotations. We further analyze the impact of language-specific and balanced pseudo-label augmentation through ablation studies. The results show that the combination of samples from multiple languages leads to better generalization and reliability. This work highlights an effective low-resource strategy for building multilingual smishing detectors, allowing a broader deployment across linguistically diverse user populations
Exploring Drivers of Technology Adoption in Public Healthcare through the TOE Lens
Get PDFThis study seeks to contribute to a deeper understanding of the adoption of technological innovations within the public health sector. By examining the interplay between technological characteristics, organizational capacity, and environmental context, the research aims to uncover how and why public healthcare organizations adopt, or resist adopting, new technologies. The Technology-Organization-Environment framework is used to understand the determinants of innovation adoption. To gain an in-depth understanding of the context of public healthcare organizations, a multiple case study design was employed. The study concludes that technology adoption in public healthcare is shaped not only by technical performance but also by organizational structures, cultural norms, and institutional dynamics. Successful adoption requires clear governance, trust, collaboration, and alignment be-tween technological solutions and the needs and values of users