Prof. Nitin Agarwal Delivers Invited Talk at the UA Little Rock Provost Leadership Council
COSMOS Research Center hosted the UA Little Rock Provost Leadership Council meeting on August 21, 2025, featuring an invited talk by Prof. Nitin Agarwal on AI-enhanced cognitive threats. He outlined how generative AI is accelerating misinformation, manipulation, and coordinated influence operations. Prof. Agarwal highlighted COSMOS’s advanced analytical, AI-driven, and network science methodologies to detect and counter these threats, emphasizing a “fight fire with fire” approach. The session encouraged cross-disciplinary collaboration to strengthen cognitive resilience and protect digital ecosystems.
Prof. Nitin Agarwal Featured in The Baltimore Sun on AI Governance in Maryland
Prof. Nitin Agarwal, Founding Director of the COSMOS Research Center at UA Little Rock, was featured in The Baltimore Sun for his expert insights on emerging AI legislation. The article highlights his perspective on balancing innovation with accountability through ethical, evidence-based AI governance. Prof. Agarwal emphasized the need for federal guiding principles paired with state-level flexibility to address diverse needs. His inclusion reflects national recognition of his leadership in AI ethics, cognitive security, and policy engagement.
COSMOS Publishes Research at NeurIPS 2025 – World’s Top AI Conference
COSMOS Research Center’s work was accepted and published at NeurIPS 2025, the world’s leading AI conference, marking a significant research milestone. The paper introduces MANGO, a novel multimodal attention-based normalizing flow framework that advances how AI systems fuse diverse data sources. Built on interpretable invertible cross-attention, MANGO enables precise, efficient, and scalable multimodal learning. Extensive evaluations demonstrate state-of-the-art performance across multiple benchmarks, reinforcing COSMOS’s leadership in transparent, high-impact AI research.
Sixteen Studies Published at Complex Networks 2025
Sixteen studies from the COSMOS Research Center, led by Prof. Nitin Agarwal, were accepted at Complex Networks 2025 in New York, highlighting COSMOS’s global leadership in computational social science and network analytics. The papers span topics such as narrative contagion, coordinated behavior, algorithmic bias, and cross-platform information diffusion, introducing novel, scalable models for analyzing influence and manipulation. Collectively, these contributions advance socio-cognitive security research and reinforce COSMOS’s mission to strengthen trust and resilience in digital information ecosystems.
Hot off the Press: Safeguarding YouTube Discussions: A Framework for Detecting Anomalous Commenter and Engagement Behaviors
A new SNAM publication, “Safeguarding YouTube Discussions,” presents a comprehensive framework for detecting anomalous commenter and engagement behaviors across 12 million users and 123 million comments. Using unsupervised learning techniques such as KDE and GMM, along with a unified anomaly scoring system, the model accurately identifies coordinated and deceptive activity. The study highlights growing influence operations, their impact on public discourse, and the need for ethical AI and interdisciplinary collaboration, reaffirming COSMOS’s leadership in advancing resilient, data-driven approaches to protecting digital ecosystems.
Research Spotlight: From Visual Persuasion to Algorithmic Bias: COSMOS Research Featured at ASONAM 2025
At ASONAM 2025 in Niagara Falls, COSMOS researchers from UA Little Rock presented ten studies, highlighting four that examine the intersection of social networks, algorithmic bias, and strategic communication. These studies explored the persuasive role of visual elements, structural dynamics in YouTube recommendation networks, algorithmic bias in YouTube Shorts, and simulation-based analyses of watch-time effects. Together, the research advances interdisciplinary insights into algorithmic influence, fairness, and cognitive security in modern information ecosystems.
New Hire: Bhushan Shelke, Data Engineer
COSMOS proudly welcomes Bhushan Shelke as its new full-time Data Engineer. With experience in data engineering, analytics, and applied data science, Bhushan specializes in building scalable pipelines, dashboards, and predictive models. He was drawn to Prof. Agarwal’s focus on addressing real-world challenges through interdisciplinary, and socially impactful research at COSMOS. Bhushan aims to deepen his expertise in large-scale social media data systems, distributed processing, and model deployment. He values the collaborative and mentoring culture at COSMOS and encourages new team members to stay curious, ask questions, and engage across domains.