Artificial Intelligence & Ethics

Global: Jurisdiction-Aware Architecture Reduces Data Exposure in Multi-Jurisdictional AI DeploymentsA team of researchers announced a new privacy‑by‑design framework on January 2026 that dynamically aligns large language model (LLM) and Internet…

Global: Study Reveals Preference-Undermining Attacks Exploit Aligned Language ModelsResearchers publishing on arXiv on January 2026 report that large language models (LLMs) trained with preference alignment—optimizing for helpful and interaction‑friendly responses—can…

Global: Temporal Consistency Learning Improves Detection of Deepfake Content, Study FindsOn 8 January 2026, researcher Prasanna Kumar and co‑authors submitted a paper to the arXiv preprint server outlining a new…

Global: Study Introduces SafeGPT Guardrails to Mitigate Enterprise LLM Data Leaks and Unethical OutputsA research team led by Pratyush Desai, along with Luoxi Tang, Yuqiao Meng, and Zhaohan Xi, released…

Global: Multi-Agent Framework Improves Control and Provenance for Generative AIResearchers announced a new multi‑agent architecture on arXiv in January 2026 aimed at enhancing user control, copyright protection, and content provenance…

Global: Online Bayesian Framework Explains In-Context Learning in Large Language ModelsResearchers have introduced a theory‑first framework that treats inference‑time adaptation in large language models (LLMs) as an instance of online…

Global: Adversarial Reasoning RAG Introduces Reasoner‑Verifier Framework for Multi‑Perspective Retrieval‑Augmented ModelsA new framework called Adversarial Reasoning RAG (ARR) was presented by a team of researchers led by Can Xu on…

Global: New Benchmark Evaluates LLMs’ Strategic Communication in Corporate Crises Researchers have introduced Crisis-Bench, a multi‑agent benchmark designed to assess large language models (LLMs) on public‑relations tasks during high‑stakes corporate…

Global: Explainable AI Perspective Highlights Role of Causal Reasoning in Scientific DiscoveryOn 9 January 2026, researchers Ricardo Vinuesa, Steven L. Brunton, and Gianmarco Mengaldo submitted a perspective paper to arXiv titled “Explainable AI: Learning from the…

Global: New Study Proposes Probabilistic Approach to Account for Ground-Truth Uncertainty in AI BenchmarksA team of researchers announced a probabilistic framework on arXiv that aims to improve the evaluation of…