News & Research
The latest AI research and news with real-world stakes — each item sourced, dated, summarized in plain English, and tagged by impact area. Every item is checked against its source before it appears.
Research
Detection, Attribution, Narration: An End-to-End Pipeline for Explainable Money Mule Identification
Yuge Zhang, Yuanxing Zhang, Yichao Jin et al.
arXiv · 2026-07-20
This paper presents a three-stage pipeline for detecting money mule accounts in financial systems, combining a LightGBM classifier built on 280 engineered features, a TreeSHAP attribution layer, and a large language model (LLM) module that converts model explanations into plain-language narratives for analysts. In live production, the system achieves an 89% yield rate versus 61% under the prior rule-based system, representing a 60% incremental improvement in adverse detection, while analyst feedback indicates the LLM-generated narratives reduce cognitive load during alert triage. The work is notable for demonstrating measurable gains in financial fraud detection at scale while also addressing explainability requirements relevant to regulated environments.
- Enterprise
- Quality assurance
- AI policy
Research
PRISM: Sensitivity-Aware PolynoMial PRuning for EffIcient Neural Network Encryption
Sahaj Majavdia, Mahdi Taheri
arXiv · 2026-07-20
PRISM introduces PSAP (Polynomial-Sensitivity-Aware Pruning), a structured pruning method designed to make neural network inference practical under homomorphic encryption (HE) while explicitly accounting for reliability. By jointly scoring filters on weight magnitude, polynomial activation sensitivity, and rotation cost, PSAP concentrates pruning in fault-tolerant regions, limiting catastrophic accuracy drops (>10 percentage points) to at most two layers compared to 5–14 layers for magnitude-pruning baselines—a worst-case vulnerability reduction of up to 29× under bit-flip injection. The method also improves computational efficiency, cutting Halevi–Shoup rotations by up to 45.2% on ResNet-32 and reducing multiplicative depth from 66 to 56 levels, enabling leveled encrypted inference without bootstrapping. These results matter for quality assurance and enterprise deployment of privacy-preserving AI, as they provide both a safe operating boundary (BER ~10⁻⁵) and a path to selective hardening of the 1.1% of parameters identified as fault-critical.
- Quality assurance
- Enterprise
Research
A Dual-Hypothesis Reasoning Framework for LLM Guardrails
Md Asiful Islam, Mihai Surdeanu
arXiv · 2026-07-20
ARBITER is a new LLM safety guardrail framework that uses dual-hypothesis reasoning—explicitly weighing both safe and unsafe interpretations of a prompt before making a safety decision—combined with a multi-component supervised fine-tuning (MC-SFT) training approach. Unlike existing methods that rely on large teacher models and full-parameter fine-tuning, ARBITER uses a cost-efficient self-generation strategy and LoRA-based parameter-efficient fine-tuning while outperforming those more expensive baselines on three safety moderation benchmarks. The framework also provides interpretable, evidence-phrase explanations for unsafe classifications, making its decisions more transparent. This matters for AI quality assurance and policy efforts, as more accurate and explainable content moderation tools are critical for responsible deployment of language models.
- Quality assurance
- AI policy
Research
ZifaMem: Structured Memory for Persona, Preference, and Emotional Continuity in AI Companions
Jingzhe Fang, Guozhi Xu, Yunfan Cui et al.
arXiv · 2026-07-20
ZifaMem is a structured memory system for AI companions that organizes dialogue into session summaries, episodic memories, and a consolidated user model to sustain emotional continuity across conversations. Tested against a deployment-realistic baseline supplying full raw dialogue history, ZifaMem raises pooled emotional-intelligence scores by 11.4% (95% CI 6.3%–17.1%) across four LLM backbones and improves persona grounding on all four, with Claude showing a +42% relative gain. Multi-turn affect context yields a +39% net preference over single-turn snapshots, while an added emotion state machine provided no measurable benefit. The system is open-sourced and its findings matter for enterprises and workforce teams building AI companion or assistant products that require sustained, personalized user relationships.
- Enterprise
- Workforce
- Quality assurance
Research
Human-in-the-Loop User Feedback Affects Perceived Accuracy and Trust, but Task Subjectivity Matters
Donald R. Honeycutt, Mahsan Nourani, Eric D. Ragan
arXiv · 2026-07-20
This paper investigates how allowing end users to provide feedback to machine learning systems (human-in-the-loop, or HITL) affects their trust in and perceived accuracy of those systems. Through three controlled experiments, the researchers found that in objective-answer contexts, providing feedback lowered both trust and perceived accuracy regardless of whether the system actually improved, while subjective feedback contexts showed no such negative bias. The study also found different patterns of trust erosion over time depending on task subjectivity. These findings are important for enterprise and workforce contexts where AI system designers must carefully consider how feedback mechanisms are structured to avoid unintended drops in user trust.
- Enterprise
- Workforce
- Quality assurance
Research
Salience Induction against Multi-Hop RAG Agents: Threat and Defense
Xingfu Zhou, Pengfei Wang, Yuan Zhou et al.
arXiv · 2026-07-20
This paper identifies a new attack surface in agentic retrieval-augmented generation (RAG) systems called 'Salience Induction,' where an adversary manipulates how facts are positioned, emphasized, or framed within documents to redirect an AI agent's reasoning — without injecting false information or explicit instructions. The authors formalize six Salience-Editing operator classes, build an iterative proposer-verifier pipeline, and introduce a benchmark called SalientWiki-MH to evaluate the threat. Experiments across five frontier model families (GPT, Claude, Gemini, DeepSeek, Qwen) and three agent architectures show that under a 30% edit budget, the attack achieves an 83.3% success rate, while the best existing baseline defense still leaves a 75.7% post-defense attack success rate. The authors also propose a lightweight 'Salience Normalization' defense that reduces attack success to 15.3%, demonstrating that truthfulness checks and instruction filtering alone are insufficient for securing agentic RAG systems.
- Quality assurance
- Enterprise
- AI policy
Research
Oracle Gap and Signal Fidelity: A Fixed-Pool Diagnostic for Test-Time Collaboration
Jie Hu
arXiv · 2026-07-20
This paper introduces a diagnostic framework for evaluating when 'test-time collaboration' techniques—such as self-consistency, best-of-N selection, critic models, and verifier pipelines—actually improve large language model reasoning. The authors decompose a verifier or selector's net gain into four measurable factors: recoverable mass, verification-signal coverage, conditional selection quality, and harm to already-correct outputs. Empirical results across LiveCodeBench, MATH Level-5, and GPQA-Diamond show that gains are bounded first by the oracle gap (how often the correct answer exists in the candidate pool) and then by signal fidelity (how well verifier verdicts align with ground truth). The framework provides a practical pre-deployment checklist—estimate oracle gap, measure coverage, signal fidelity, and harm—before investing in collaboration infrastructure, which matters for enterprise and quality-assurance teams deploying LLM pipelines.
- Enterprise
- Quality assurance
Research
Can AI Agents Really Complete RTL-to-GDS? Lessons from Benchmarking Tool-Interactive EDA Workflows
Jinyuan Deng, Zhengrui Chen, Xufeng Wei et al.
arXiv · 2026-07-20
FluxBench is a systematic benchmark that evaluates AI agent systems on complete end-to-end electronic design automation (EDA) workflows, from RTL code generation through placement and routing (RTL-to-GDS), using unified prompts and tool environments. The study finds that even agents built on the same foundation model can show performance gaps of up to 86.27% depending on their system architecture, and that cost-efficiency (measured by the new Token ROI metric) can vary by as much as 105.92× among systems with comparable task performance. In an RTL-to-GDS case study using PicoRV32, their FluxEDA system achieves an end-to-end score of up to 97.94, outperforming Claude Code with domain-specific EDA skills by up to 8.39×. The results demonstrate that domain-specific skills alone are insufficient for large-scale EDA automation, and that both agent system design and underlying model capability are critical factors—a finding with significant implications for how AI tools are architected and deployed in industrial chip design workflows.
- Enterprise
- Quality assurance
Research
Thinking in Video: Can Video Generators Really Reason About the Real World?
Yongheng Zhang, Guang Yang, Ruihan Hou et al.
arXiv · 2026-07-20
This paper introduces the Causal-Generative Dual-Judge (CGDJ), a framework that evaluates whether video generative models can genuinely reason about real-world causal dynamics or merely reproduce memorized visual patterns. The authors define 'Thinking in Video' as a paradigm where video serves as a medium for causal reasoning, then apply CGDJ to both open- and closed-source generators, finding a clear Perception-Prediction Gap: open-source models produce plausible-looking videos despite near-zero explicit causal understanding, while closed-source systems show stronger but still limited alignment between reasoning and generation. A key finding is audio-visual misalignment, where models can verbalize correct causal logic more reliably than they can render it visually, undermining claims that current systems function as true 'world simulators.' These results matter for quality assurance and enterprise deployment of video AI, as they reveal that perceptual fidelity is insufficient evidence of causal understanding.
- Quality assurance
- Enterprise
Research
Retrieval-Augmented Interpretable Learning: Towards Task-Specific Zero-Shot Models in Healthcare
Sazan Mahbub, Caleb Ellington, Zhiyuan Li et al.
arXiv · 2026-07-20
RAIL (Retrieval-Augmented Interpretable Learning) is a probabilistic meta-learning framework that generates interpretable, task-specific clinical prediction models from natural-language task descriptions and a memory of previously learned predictors—without requiring task-specific training data. In zero-shot settings it achieves 73.4% accuracy on long-tailed clinical procedure prediction tasks, and maintains 73.2% accuracy with only 2–4 examples, where conventional supervised models perform near chance. The framework provides uncertainty estimates over retrieval, model coefficients, and predictions, allowing unreliable outputs to be flagged for clinical review rather than acted upon automatically. These properties make RAIL particularly relevant for healthcare environments where new clinical targets arise frequently, interpretability is required, and human oversight is essential.
- Quality assurance
- Enterprise
- AI policy
Research
DecoyFace: Beyond Obfuscation via Controllable and Imperceptible Identity Misdirection for Privacy-Preserving Face Recognition
Zhihan Ren, Lijun He, Xinyao Wang et al.
arXiv · 2026-07-20
DecoyFace is a privacy-preserving face recognition framework that addresses a key weakness in split face recognition systems: intermediate features sent to a server can be inverted to reconstruct the original identity. Rather than producing visibly degraded reconstructions that signal the existence of protection, DecoyFace injects plausible but incorrect 'decoy' identity cues into the reconstruction-sensitive part of the feature representation, steering attackers toward a false identity while retaining enough information for accurate recognition via an authorized server-side canonicalization module. Experiments on the LFW dataset show identity leakage drops to 2.93% under U-Net attacks and 0.74% under Flow-Matching attacks, with over 99.78% face validity, meaning the decoy reconstructions remain visually convincing and imperceptible as fakes. This work matters for enterprise deployments and quality-assurance of AI systems where biometric data privacy must be maintained without alerting adversaries to the presence of protection mechanisms.
- Enterprise
- Quality assurance
- AI policy
Research
Pailitao-MMSearch: Building Native E-Commerce Multimodal Search Foundation
Xiaohan Ye, Xu Chen, Zihan Gong et al.
arXiv · 2026-07-20
Pailitao-MMSearch introduces a native multimodal search foundation model for e-commerce that unifies text, image, and mixed-intent queries within a single system, replacing the fragmented approach of separate specialist models. The system incorporates three innovations—Hybrid Semantic ID (HybSID), a two-stage continual pre-training strategy, and a hybrid reasoning post-training pipeline—built on the Qwen base model and deployed on Taobao's Pailitao platform. Online A/B testing results show up to +13.61% improvement in Gross Merchandise Volume (GMV) and +8.21% in transaction volume compared to traditional multimodal search pipelines, demonstrating meaningful commercial gains from domain-adapted large language models in e-commerce search.
- Enterprise
Research
Multilingual Sentence Embeddings for Linguistic-Integrated Reliability Audit
Ummugul Bezirhan, Ji Yoon Jung, Matthias von Davier
arXiv · 2026-07-20
This paper investigates whether multilingual sentence embeddings can substitute for translated English text in Linguistic-Integrated Reliability Auditing (LiRA) for educational assessment. Tested across 11 PIRLS constructed-response items and three embedding models, native-language embeddings closely reproduced reliability estimates previously derived from translated input. The approach also recovered responses that had been excluded due to translation failures, without meaningfully changing reliability outcomes. This matters for quality assurance in multilingual assessment systems, as it suggests translation can be bypassed without sacrificing scoring consistency.
- Quality assurance
- Certifications
Research
How Reliable Are Multimodal Signals of Conversational State? Evidence from Remote Dyadic Collaborative Tasks
Tahiya Chowdhury
arXiv · 2026-07-20
This paper evaluates how reliably multimodal signals—linguistic, acoustic, and interactional features—can measure conversational states like cognitive load and conversational power in remote dyadic collaborative tasks. Using a three-dimensional framework (predictive accuracy, cross-task generalizability, and test-retest reliability) applied to the AVCAffe dataset of 53 dyads across 9 tasks, the authors find that linguistic features predict cognitive load best but generalize poorly across tasks due to vocabulary sensitivity, while acoustic reliability collapses to near-zero once speaker identity is controlled. Only interaction features, such as floor dominance, provide genuinely reliable signals for measuring within-dyad cognitive load asymmetry, and classifying power roles remained near chance across all conditions. The findings argue that speaker normalization and multi-dimensional evaluation are prerequisites for robust multimodal feature selection in conversational AI systems.
- Quality assurance
- Enterprise
Research
Calibrating Semantic Uncertainty from Observable Language-Model Probabilities
Matthew F. Dixon
arXiv · 2026-07-20
This paper addresses the gap between a language model's raw token probabilities and the meaningful uncertainty professionals need for decisions about states like diagnoses or economic conditions. The authors introduce a 'semantic map'—a semiparametric statistical framework that translates word-level model outputs into calibrated posterior probabilities over declared semantic states, with testable acceptance conditions and auditable uncertainty bounds. Validated on Federal Reserve financial text and controlled simulations, their approach outperforms printed model confidence scores, maintains stability under paraphrasing, and responds correctly to new evidence. The work has direct implications for enterprise and quality-assurance settings, offering a principled method to audit and certify when AI-generated language outputs reliably reflect application-relevant meaning.
- Enterprise
- Quality assurance
- Certifications
- AI policy
Research
The Dependency Dilemma: How Machine Learning Decision Aids can Undermine Skill Growth
Kevin Bauer, Michael Nofer, Benjamin Henrich et al.
Business & Information Systems Engineering · 2026-07-20
This study uses a novel experimental design to causally identify how relying on ML decision aids affects employees' decision-making skill development. The findings show that reliance on ML predictions hinders the growth of critical decision-making skills, leading to significant performance drops when the system becomes unavailable, with the degree of trust in the system amplifying the skill deficit. The research highlights the need for organizations to thoughtfully balance ML aid reliance with human skill retention, particularly to manage risks from system disruptions.
- Workforce
- Enterprise
Research
Pengaturan Kecerdasan Artifisial dan Big Data dalam Perspektif Hukum serta Tantangan dan Peran Pemerintah di Era Digital
Sika Aprilia
Jurnal Ilmu Hukum Humaniora dan Politik · 2026-07-20
This Indonesian legal study examines the regulatory landscape for Artificial Intelligence and Big Data through a normative legal research approach, using statutory and conceptual analysis. The research finds that Indonesia's current regulations on AI and Big Data are partial and not comprehensively integrated, creating gaps in areas such as personal data protection, information security, algorithmic transparency, and potential technology misuse. The paper argues that the government needs to strengthen policy through a 'smart regulation' approach that is adaptive and responsive to technological developments to ensure these technologies align with legal principles and public interest.
- AI policy
Research
Research on credit portfolio optimization decisions under digital risk control: an integrated framework based on explainable AI and hybrid intelligence
Sujuan Pan, Yilin Wu, X K HUANG
Humanities and Social Sciences Communications · 2026-07-20
This study proposes an integrated framework for digital credit governance that links default-risk prediction, risk quantification, and portfolio optimization into an auditable decision chain. It combines XGBoost-SHAP for explainable prediction, EWM-TOPSIS for risk evaluation, and a hybrid simulated annealing-neighborhood algorithm for portfolio allocation, achieving an AUC of 0.851 while improving RAROC from 12.5% to 17.5% and reducing 95% CVaR from 3.10M to 2.53M relative to a rules-based baseline. The framework's primary contribution is making credit-allocation decisions more auditable and open to supervisory scrutiny, offering a transferable design logic for lending institutions seeking interpretable and reviewable AI-driven decision systems.
- Enterprise
- AI policy
Research
La inteligencia artificial en el Perú: análisis normativo en torno al ciclo de vida del dato personal y de los sistemas de IA
Alexandra Espinoza Montero
Derecho & Sociedad · 2026-07-20
This article uses comparative normative analysis to examine how AI system lifecycles intersect with personal data lifecycles under Peruvian law, mapping privacy and data protection obligations at each stage. It integrates international frameworks—including the EU AI Act, GDPR, OECD guidelines, and ISO standards—with Peru's national regulatory context, and analyzes the specific responsibilities of actors such as providers, deployers, importers, and data processors. The paper emphasizes principles like privacy by design, algorithmic audits, and impact assessments to prevent risks such as discrimination, bias, and opacity. It concludes that a flexible, principles-based regulatory approach complemented by technical instruments like ISO standards is more effective than rigid regulation for promoting responsible AI implementation in Peru.
- AI policy
- Certifications
Research
ChannelGuard: Safe Models Do Not Compose into Safe Multi-Agent Systems
Elias Hossain, Md Mehedi Hasan Nipu, Fatema Tuj Johora Faria et al.
arXiv (Cornell University) · 2026-07-20
ChannelGuard addresses a critical blind spot in multi-agent LLM security: while individual models may appear safe, the inter-agent communication channels between planner, worker, verifier, and synthesizer components are unmonitored and exploitable. The paper shows through a 2,100-trace evaluation that apparent pipeline safety often depends silently on provider-side cloud filters rather than application-layer defenses, a dependence hidden by outcome-only reporting. ChannelGuard proposes a training-free defense-in-depth framework that places information-bottleneck gates on every inter-agent channel, achieving perfect blocking of Tool Poisoning attacks (30/30) consistently across multiple model backends and halving Prompt Injection success rates, while preserving task accuracy. The work highlights that safe individual components do not automatically compose into safe multi-agent systems, with implications for how enterprise deployments and quality assurance processes evaluate AI pipeline security.
- Quality assurance
- Enterprise
Research
Advancing Automatic Recognition through Digital Transformation:Balancing Automation with a Human-Centric Approach
Luca Ferranti, Chiara Finocchietti, Serena Spitalieri
Universitas - studi e documentazione di vita universitaria · 2026-07-20
This paper examines how digital transformation (DT) of academic credential recognition can advance Automatic Recognition (AR) without replacing human judgment with full technological automation. Analyzing binding regulations including the EU AI Act—which classifies AI systems used in education and qualification evaluation as high-risk—alongside the Lisbon Recognition Convention and other international frameworks, the authors argue that AR must be distinguished from total automation. Using normative document analysis and the CIMEA Italian ENIC-NARIC center as a case study, the study identifies six pillars for a compliant DT strategy: Human Oversight and Accountability, Human-Centric Design, Ethics-by-Design and Quality-by-Design, Robust Data Governance and Privacy, Transparency and Explainability, and AI Literacy/Upskilling for staff.
- AI policy
- Certifications
Research
A GOVERNANCE MODEL FOR AGENTIC AI USE IN BANKING SYSTEMS
International Research Journal of Modernization in Engineering Technology and Science · 2026-07-20
This paper proposes a risk-aware governance framework specifically designed for agentic AI systems deployed in banking, addressing gaps in traditional model-centric AI governance. The framework organizes controls across policy governance, planning oversight, tool authorization, runtime monitoring, human supervision, and audit evidence management, and introduces a five-level autonomy model ranging from advisory support to restricted high-autonomy execution. Banking use cases such as fraud investigation, AML operations, onboarding, and compliance are analyzed to show how governance intensity should scale with operational risk. The work shifts AI governance discourse from model validation toward bounded autonomy, runtime control, and audit-ready traceability in regulated financial environments.
- AI policy
- Enterprise
Research
Who Can Stop a Frontier Model? Testing, Release Authority, and the Limits of the FINRA and FAA Analogies
Peter Bell
Zenodo (CERN European Organization for Nuclear Research) · 2026-07-20
This working paper analyzes the institutional design of frontier AI model testing and release authority by comparing two prominent proposals—a FINRA-like standards body (proposed by Demis Hassabis) and an FAA-like mandatory evaluation regime (proposed by Dario Amodei)—against the actual legal machinery of those analogues. The paper finds that while standards development, inspection, testing, and certain certification activities can be delegated to private or hybrid bodies, functions such as compulsory access, release prohibition, enforcement, sanctions, and judicial review must rest on an explicit public-law authority chain. It introduces a reusable diagnostic framework for evaluating which actors may lawfully perform which governance functions in a frontier-model release regime. The analysis is presented as an institutional-design inference rather than an empirical assessment of whether either proposed regime would be effective.
- AI policy
- Certifications
Research
Algorithmic fairness and bias mitigation in financial artificial intelligence: scoping review
Marcelo Wecchi, Lilian Berton
AI and Ethics · 2026-07-20
This scoping review systematically maps how academic literature addresses algorithmic fairness and bias mitigation in financial AI, drawing on 99 articles screened from over 17,000 records. The authors find that credit risk and credit scoring dominate the research landscape, while standardized fairness metrics remain lacking and intersectionality and causal approaches to sensitive data are underexplored. The review identifies methodological gaps and calls for more consistent evaluation frameworks to support both future research and policy development in AI-driven financial decision-making.
- AI policy
- Enterprise
Research
Artificial Intelligence Readiness in Clinical Trial Operations: A Narrative Review and Site-Level Governance Framework
Simona Wójcik, Anna Rulkiewicz, Justyna Domienik‐Karłowicz
Preprints.org · 2026-07-20
This narrative review examines how AI is being applied across clinical trial operations and finds that no current use case reaches high evidence maturity—even the strongest area, patient-trial matching and eligibility assessment, was evaluated only retrospectively or in simulated settings rather than in live trials. The authors identify recurring risks including hallucination, automation bias, model drift, and unclear accountability, and propose a site-level governance framework that links evidence maturity, AI autonomy, trial impact, and site implementation capacity. The paper argues that AI readiness must be assessed at the level of the AI-enabled workflow, not just the model, and that safe adoption requires context-specific validation, human accountability, auditability, and alignment with Good Clinical Practice.
- Quality assurance
- AI policy