AI Frontier Institute

News & Research

The latest AI research and news with real-world stakes. Each item is sourced, dated and summarized in plain English, tagged by impact area where one fits, and its summary is checked against the text it was written from.

• ResearchSystems2026-07-10Enterprise

  Artificial Intelligence and Labor Productivity in Construction: A Comparative Systems Analysis Across European Economies · Claudiu George Bocean, Adriana Scrioșteanu, Sorina Gîrboveanu et al.

  This study examines how AI adoption affects labor productivity in the construction sectors of EU countries using 2023–2024 data. Employing multivariate log-linear regressions and cluster analysis, the researchers find that construction output is the primary driver of labor productivity, while AI adoption shows a small, negative association with productivity, interpreted as short-term adjustment costs during early digital transformation. Cluster analysis reveals diverse country profiles in AI use, productivity, and labor intensity, suggesting that blanket digital adoption policies may be insufficient. The authors conclude that gradual, organization- and skills-focused change management is needed to realize long-term productivity gains from AI in construction.
• ResearchWorld Journal of Advanced Research and Reviews2026-07-10Quality assurance · AI policy

  A definitive forecast on the regulation of generative artificial intelligence for the years 2025 to 2035 · Jimmy Kinyonyi Bagonza, Mathias Ndungu

  This paper presents a qualitative, scenario-based forecast of how generative AI may be regulated globally between 2025 and 2035, identifying two critical uncertainties: the level of international regulatory coordination and the pace of AI safety and interpretability advances. These uncertainties underpin two contrasting scenarios—Regulated Convergence, featuring strengthened international cooperation and technical verification mechanisms, and Fragmented Acceleration, characterized by geopolitical competition and limited safety progress. The study finds that effective governance by 2035 will require sustained institutional investment, adaptive regulatory frameworks, and greater international collaboration, and warns that without these conditions, fragmented governance and increased societal risks from frontier AI systems are the likely outcome. The paper is directly relevant to how policymakers, certification bodies, and enterprises should anticipate and respond to evolving AI governance regimes.
• ResearchManager Zdravookhranenia2026-07-10Enterprise

  The process of creating an artificial intelligence-based agent for solving management ophthalmology tasks. · A. I. Bursov, A. V. Belogurova

  This paper describes the development of an AI agent designed to automate competitive intelligence for private ophthalmology clinics, replacing manual monitoring of competitor websites and price lists. Built on Python, LangChain, retrieval-augmented generation (RAG), and locally deployed large language models (Qwen 2.5 and Llama 3.1), the agent crawls clinic websites, structures data, and generates comparative reports in response to management queries. Testing showed that typical query execution time dropped from approximately 2 hours of manual work to under 12 minutes (as low as 1.5–2 minutes with cloud models), demonstrating significant labor cost reduction. The authors argue the framework is adaptable to medical organizations beyond ophthalmology and supports more informed, timely managerial decision-making.
• ResearchInternational Journal of Computer Information Systems and Industrial Management Applications2026-07-10AI policy · Education

  The AI-Native University: A Conceptual Framework and Maturity Model for AI Transformation in Higher Education · Virendra Gawande, Sarika V. Gawande

  This paper introduces the 'AI-Native University' (AINU) concept, a framework proposing that higher education institutions treat AI not as isolated tools but as a core institutional capability embedded across teaching, learning, assessment, research, and administration. The authors develop a maturity model to help universities assess their current AI adoption level and plan progression toward integrated, institution-wide AI transformation, supported by governance, ethics, quality assurance, and faculty development. The framework is intended to guide university leaders, policymakers, and researchers in benchmarking AI readiness and informing policy development. The work highlights the growing importance of human-AI collaboration and ethical oversight at the institutional level.
• ResearchSustainability2026-07-10Enterprise

  From AI Use to Sustainable Value Creation Through Entrepreneurial Reconfiguration and Business Model Innovation in SMEs: Evidence from an Emerging Economy · Alexander Sánchez-Rodríguez, Jesús Rodríguez-Flores, Reyner Pérez-Campdesuñer et al.

  This study examines how AI use translates into sustainable value creation for small and medium-sized enterprises (SMEs) in Ecuador, using survey data from 385 firms across four sectors. Using PLS-SEM, the researchers found positive associations between AI use, entrepreneurial reconfiguration capability, business model innovation, and sustainable value creation, suggesting that AI adoption alone is insufficient without developing the organizational capabilities to connect AI with new business models. The findings are particularly relevant to emerging-economy contexts where AI adoption is often fragmented. The study shifts focus from mere AI adoption to AI-enabled entrepreneurial transformation as the pathway to measurable business impact.
• ResearchZenodo (CERN European Organization for Nuclear Research)2026-07-10Quality assurance · Certifications · +3

  The Admissibility Threshold: A Sector-Specific Certification Standard for High-Stakes AI, with Health as the First Mandatory Domain (Version 2) · Siddiqui Jameel Ahmed

  This paper introduces 'Domain Admissibility,' a sector-specific certification framework requiring AI systems to prove fitness for a particular high-stakes domain before deployment, rather than satisfying only generic horizontal governance standards. The authors argue that uniform AI governance is inadequate when error costs vary drastically across domains—most critically in healthcare, where mistakes can be irreversible and fatal. They define four testable pillars of admissibility, propose a certification threshold for consequential settings, and make the case that health should be the first domain where this threshold becomes mandatory. The work positions itself as a foundational constitutional framework on which future certification institutions could be built, addressing a gap left by existing medical-device regulation that does not cover general-purpose AI entering clinical settings.
• ResearchWorld Journal of Advanced Research and Reviews2026-07-10Enterprise · Quality assurance · +2

  Runtime assurance for enterprise agentic AI systems: A policy-gated control model with quantitative autonomy-risk scoring · Kwan Hong Tan

  This paper presents a Runtime Assurance Architecture (RAA) for enterprise agentic AI systems that introduces a quantitative Autonomy-Risk Exposure (ARE) score to govern when AI agents can act autonomously, require sandboxing, need human approval, or must be blocked. Evaluated across 2,000 simulated enterprise agent episodes, the full RAA configuration reduced mean ARE scores by 31.5%, cut policy-conflicting actions from 9.8% to 4.6%, eliminated unsupervised high-risk pass-throughs, and improved audit evidence coverage from 0.61 to 0.89—all with a mean latency overhead of 95 ms. The findings are directly relevant to organizations deploying agentic AI in regulated or high-consequence workflows, offering a practical reference architecture and policy decision algorithm grounded in AI risk-management standards and security guidance.
• ResearcharXiv2026-07-09Quality assurance

  The Patchwork Problem in LLM-Generated Code · Viraaji Mothukuri, Reza M. Parizi

  This paper identifies and formalizes what it calls the 'patchwork problem': LLM-generated code that compiles and passes tests but is globally incoherent due to structural failures such as missing configuration keys, nonexistent packages, or omitted authentication guards. The authors model structural coherence as consistency invariants over graph representations of repository artifacts and introduce an eight-category failure taxonomy, distinguishing defects unique to LLM generation from those merely amplified by it. Their hybrid verification framework combines mature static analysis tools with purpose-built detectors, and empirical evaluation shows the vast majority of these structural failures evade type checking, testing, and existing SAST tools entirely. External validation on real-world AI-generated repositories confirms these failures are widespread wherever LLMs write code with minimal human oversight, posing a significant and growing risk to software quality.
• ResearcharXiv2026-07-09Enterprise · Algorithms & Automated Decisions

  Eluna: An Agentic LLM System for Automating Warehouse Operations with Reasoning and Task Execution · Ning Liu, Kalle Kujanpää, Zhaoxuan Zhu et al.

  Eluna is a production-deployed agentic AI system designed to automate warehouse operations by encoding Standard Operating Procedures (SOPs) as directed acyclic graphs within a multi-agent framework that enforces procedural compliance and handles complex, multi-system decision logic under strict time constraints. The system uses asymmetric episodic distillation, where a strong teacher model is refined through episodic error memories and a smaller student model is fine-tuned on corrected trajectories, allowing smaller models to match or exceed larger off-the-shelf baselines without inference-time overhead. On a 13-task benchmark and two production applications, Eluna's fine-tuned models achieve 94% expert agreement on a ticket processing application, demonstrating reliable SOP execution at scale. This work is significant for enterprise warehouse automation, showing that purpose-built agentic frameworks can outperform general large language models on complex operational tasks.
• ResearcharXiv2026-07-09Quality assurance · Health

  Trivial Prompt Reframing Bypasses Safety Guardrails in Googleś MedGemma-4B · Avi-ad Avraam Buskila

  This paper evaluates the safety guardrails of MedGemma-4B-it, Google's open-weight medical language model, by testing whether simple, non-technical prompt reformulations can bypass restrictions the model card prohibits—such as recommending drug dosages, issuing diagnoses, or advising patients to skip emergency care. Using a benchmark of 4,500 generated responses across five guarded behaviors, six attack styles, and three judge methods, the authors find an overall Attack Success Rate of 38.0%, with reframing a question as a 'medical board exam' item raising success rates to 53.1% and appeals to claimed doctor authority reaching 43.7%. The drug-interaction guardrail proved nearly absent at 83.2% attack success, while emergency-deferral was more robust at 4.7% but still breachable via authority framing. The findings demonstrate a substantial gap between model card intent and actual robustness, motivating stronger deployment-time safety measures for open medical AI models.
• ResearcharXiv2026-07-09Quality assurance · Education

  L2-Bench: An Evaluation Benchmark for Measuring LLM Capabilities in Second Language Education · James Edgell, Wm. Matthew Kennedy, Ben Knight et al.

  L2-Bench introduces an open-source benchmark of 1,000+ task-response pairs designed to evaluate large language models on their ability to apply second language (L2) education principles, not merely recall them. The benchmark includes a taxonomy of 12 competencies and 31 subcompetencies validated by 200+ expert practitioners, alongside a rubric-based evaluation methodology intended to generalize to other open-ended, qualitative educational domains. Among tested models, Claude Opus 4.7 performs best overall at 85.5%, though performance drops notably on harder tasks. The work equips education stakeholders with more rigorous tools for making informed decisions about adopting, using, and governing AI-powered educational systems.
• ResearcharXiv2026-07-09Education

  Using AI-based Learning Assistants in Higher Education: A Large-Scale Descriptive Analysis · Kristina Schaaff, Quintus Stierstorfer, Valerie Hekkel

  This study analyzes real-world usage of an AI-based learning assistant called Syntea across 77,543 distance-learning students in higher education, using objective log data rather than self-reported surveys. The researchers examine how usage patterns vary by gender, age group, study cluster, degree, and study mode, finding that Syntea is already embedded in many learners' study routines but that usage differs across demographic and structural contexts. The work addresses a gap in prior research, which has largely relied on small samples and self-reported data, by providing large-scale empirical evidence of actual educational chatbot behavior. These findings offer a foundation for improving AI-based learning support tools in higher education settings.
• ResearcharXiv2026-07-09Quality assurance · AI policy

  Trusting sovereign language models as scientific instruments: evidence from Portugal's AMALIA · Manuel Pita

  This paper audits whether nationally sovereign language models can be trusted as valid scientific measurement instruments, using Portugal's publicly funded AMALIA model as a test case. The authors introduce the 'recovery gap' metric, which checks whether a model's coding performance can actually be attributed to the theoretical construct it is supposed to measure—rather than surface-level correlates—by decomposing a codebook into theory-defined clauses and recombining them. Applied to moral foundation coding in European Portuguese, AMALIA achieves competitive agreement with human coders but only about half of its coding performance on the authority foundation can be traced back to the underlying theory, while a larger multilingual model closes this gap. The findings argue that public ownership and linguistic specialization earn operational trust but not epistemic trust, and that the proposed audit method is inexpensive and portable across models, languages, and tasks.
• ResearcharXiv2026-07-09Quality assurance · AI policy · +2

  SolarChain-Eval: A Physics-Constrained Benchmark for Trustworthy Economic Agents in Decentralized Energy Markets · Shilin Ou, Yifan Xu, Luyao Zhang

  SolarChain-Eval introduces a physics-constrained benchmark for evaluating autonomous AI agents operating in decentralized energy markets, framing market governance as a Markov Decision Process assessed across dimensions such as market utility, physical safety, slippage, action smoothness, spatial fairness, and auditability. The benchmark incorporates an LLM-based Planner/Auditor layer that defines action bounds, reviews high-risk decisions, and logs all interventions with structured audit traces. Experiments comparing static, random, myopic, RL, and RL+LLM policies reveal a clear utility-safety trade-off: RL agents boost market utility but can still produce unsafe behavior, and removing physics penalties causes reward-maximizing agents to exploit invalid generation data and inflate artificial liquidity. The findings demonstrate that trustworthy agentic AI evaluation in cyber-physical settings requires both hard physical constraints and transparent intervention records, with implications for quality assurance and policy governance of autonomous economic systems.
• ResearcharXiv2026-07-09Quality assurance · Health · +1

  The complexities of patient-centred conversational artificial intelligence · João Matos, Olivia Buege, Donny Cheung et al.

  This paper analyzes 2,053 real patient-chatbot conversations to show that communication patterns and emotional expression vary widely among users—far beyond what cooperative, idealized simulated patients capture. The researchers built a patient simulator modeling clinical content, emotional state, conversational strategy, and communication style, producing conversations so realistic that human graders achieved only 55% accuracy distinguishing them from real ones. Testing four large language models across 1,164 clinician-graded cases using five distinct patient personae, they found that communication style significantly alters triage outcomes. The findings warn that health chatbots designed for idealized interactions risk underperforming and amplifying health disparities when deployed with real, diverse patients.
• ResearcharXiv2026-07-09Quality assurance · Health · +1

  Towards Precision Therapy in Hepatocellular Carcinoma: A Clinical-Reasoning LLM for Risk Stratification and Treatment Guidance · Peng Cui, Jitao Wang, Siyan Xue et al.

  HCC-STAR is a large language model designed to support clinical decision-making in hepatocellular carcinoma (HCC) by reading electronic medical record narratives and jointly producing risk-based staging, ranked treatment recommendations, and individualized survival estimates. Trained on approximately 30,000 SEER-derived cases expanded into EMR-style narratives, the model was evaluated on a multi-center cohort of 6,668 patients across 12 hospitals in China, achieving state-of-the-art performance compared to clinical guidelines and leading models including GPT-5 and Gemini-2.5 Pro. Hypothetical overall-survival analysis showed a median survival of 51 months under HCC-STAR recommendations versus 29 and 32 months under BCLC and CNLC guidelines respectively. Blinded hepatobiliary specialists rated HCC-STAR's reasoning as trustworthy, and the model outperformed resident and attending physicians in treatment accuracy while helping clinicians make more accurate decisions faster.
• ResearcharXiv2026-07-09Enterprise

  The Context Access Divide: Interaction-Level Architecture as a Complementary Dimension of Agentic Inequality · Masahiro Fujita

  This paper introduces the 'Context Access Divide' (CAD) as a new dimension of AI inequality operating at the individual interaction level, complementing Sharp et al.'s (2025) framework of agentic inequality across availability, quality, and quantity. The authors argue that two users with nominally equivalent AI agent access can experience qualitatively different utility depending on whether the system autonomously retrieves relevant context (Dynamic Context Retrieval) or requires users to manually attach documents at each query (Manual Attachment). Using a probabilistic model grounded in the fan effect literature from cognitive psychology, they demonstrate that manual context attachment leads to combinatorial collapse in task-success probability as a user's knowledge corpus grows and tasks become more conjunctive, while dynamic retrieval architectures avoid this collapse. The paper analyzes the technical underpinnings of this divide in Model Context Protocol (MCP) and retrieval-augmented generation (RAG) architectures, and examines implications for knowledge-work stratification and AI platform governance.
• ResearcharXiv2026-07-09Quality assurance · AI policy

  Two Axes of LLM Abstention: Answer Correctness and Question Answerability · Benedikt J. Wagner

  This paper investigates why large language models (LLMs) fail to appropriately refuse both wrong answers and unanswerable or false-premise questions. The authors find that these are two distinct axes: standard confidence scores can detect when a model will answer incorrectly, but are nearly blind to whether a question is actually answerable, and this blind spot does not improve with model scale. A hidden-state linear probe fills the gap for answerability detection (reaching 0.69–0.77 AUROC on false-premise questions), and combining both signals into a two-axis calibrated policy dramatically outperforms single-threshold approaches—achieving 0.75 coverage of correct answers versus 0.31 for a single threshold. These findings matter for quality assurance and policy around LLM deployment, as they show that reliable abstention requires separately certifying correctness and answerability budgets rather than relying on a single confidence score.
• Newsnist.gov2026-07-09Enterprise · Quality assurance · +1

  New Fabric Test Material Could Help Strengthen Domestic Supply Chain for Textiles and Clothing

  NIST News reports that researchers at the National Institute of Standards and Technology have developed a new Research Grade Test Material (RGTM 10279) consisting of five fabric squares made from different fibers, designed to help the textile industry validate and improve methods for identifying and sorting textiles. The material is intended to support AI-enabled sorting technologies, which the agency says have not yet been exhaustively tested for accuracy in fiber identification. NIST is distributing the free test material to labs and manufacturers through July 30, 2026, in exchange for measurement feedback, with the goal of ultimately developing a more robust reference standard that meets real-world industry needs. Researchers note the material could also help verify fabric composition for brands and potentially support quality control across the domestic textile supply chain.
• ResearcharXiv2026-07-09Quality assurance · AI policy

  Reverse Engineering Compliance: A Dual-Graph Verification Framework for Auditing Legacy IT Security Concepts · Lea Roxanne Muth, Marian Margraf

  This paper introduces ASSERT, a framework for auditing legacy IT security concept documents by extracting them into formal document graphs and comparing them against a verified reference graph using a five-class graph difference method. The framework exports schema-valid OSCAL artifacts to support auditable, machine-readable compliance evidence, addressing a gap left by prior work that focused on generating new security concepts rather than verifying existing ones. Evaluated on the BSI's RecPlast dataset, ASSERT reveals a trade-off between discovering undocumented infrastructure entities and enforcing a strict schema, making document-infrastructure inconsistencies measurable. This work is directly relevant to organizations facing NIS-2 Directive compliance obligations and to BSI's ongoing Grundschutz++ initiative.
• ResearcharXiv2026-07-09Quality assurance · National Security & Defense

  From Legacy Documentation to OSCAL: An MCP-Based Agent Pipeline for Threat-Informed Continuous Compliance in Critical Infrastructure · Lea Roxanne Muth, Marian Margraf

  This paper presents a multi-agent AI pipeline that converts natural-language descriptions of critical infrastructure systems into structured, audit-ready compliance artifacts in the NIST OSCAL format, without requiring active network scanning of sensitive operational technology environments. The pipeline grounds LLM reasoning in authoritative threat-intelligence sources via a Model Context Protocol (MCP) architecture, reducing hallucinated vulnerabilities and attack paths. In a synthetic water utility scenario, the system achieves 0.90 CVE recall and perfect D3FEND recall, producing schema-valid OSCAL System Security Plans and Security Assessment Reports. The key finding is that grounding shifts errors to the asset-extraction phase, making remaining risks visible and suitable for efficient manual review rather than eliminating errors entirely.
• ResearcharXiv2026-07-09Quality assurance

  Psychological Competence as a Missing Dimension in AI Evaluation · Marcos Economides, Paul M. Sacher, Samuel Salzer et al.

  This paper argues that current AI evaluation frameworks are incomplete because they focus on technical metrics like accuracy and robustness while ignoring how AI systems psychologically affect the users they interact with. The authors introduce 'psychological competence' as a new evaluation dimension, defined as an AI system's capacity to support user cognition, emotional interpretation, and behavioral decision-making in contextually appropriate ways. They outline a conceptual framework covering interaction properties such as framing, tone, perceived authority, and uncertainty handling, and propose assessment approaches including scenario-based probes and structured human evaluation. The work has direct implications for model providers, deploying organizations, and regulators seeking to understand the real-world effects of AI systems used as advisors, coaches, tutors, and companions.
• ResearcharXiv2026-07-09Quality assurance · Certifications · +2

  MentalHospital: A Virtual Environment for Evaluating Psychiatric Clinical Encounters · Yuming Yang, Xiao Sun, Yuanwei Zou et al.

  MentalHospital is a virtual evaluation environment that benchmarks large language models on complete psychiatric clinical encounters, following the S.O.A.P. workflow across 1,193 de-identified EHR cases spanning all major ICD-11 categories and 76 disorders. The framework uses a dual-track assessment protocol combining objective EHR-derived references with subjective clinical process quality, supported by MentalEval, a set of five domain-specific AI evaluators achieving an average quadratic weighted kappa of 0.944 against expert judgment. Benchmarking results show that even the strongest LLM trails clinicians by 37.28 percentage points in objective psychiatric competence, with mental status assessment identified as a key bottleneck. These findings highlight significant gaps between current AI capabilities and clinical standards in psychiatry, with implications for the quality assurance and certification of AI systems in high-stakes healthcare settings.
• ResearcharXiv2026-07-09Quality assurance

  Prismata: Confining Cross-Site Prompt Injection in Web Agents · Corban Villa, Alp Eren Ozdarendeli, Sijun Tan et al.

  Prismata is a defense system designed to protect autonomous web agents from cross-site prompt injection attacks, where malicious third-party or user-generated content on a webpage hijacks an agent by masquerading as instructions. The system enforces contextual least privilege by dynamically deriving trust labels for page content and using structural confinement — inspired by classical integrity models — to ensure labeling errors only reduce privilege rather than escalate it. Mechanical confinement then redacts suspicious content and restricts agent capabilities accordingly, requiring no developer annotations and thus supporting arbitrary websites. Evaluated against recent published web agent attacks including adaptive variants, Prismata substantially reduces attack success while preserving the agent's ability to complete legitimate tasks.
• ResearcharXiv2026-07-09Quality assurance

  Gauge dependence and structured-output corruption in sign-branched repetition penalties: measurements across models, inference stacks, and alternative repetition controls · Peter Hollows

  This paper identifies a fundamental flaw in the multiplicative repetition penalty used across major LLM inference engines (HuggingFace, vLLM, llama.cpp, and others): the penalty branches on the sign of raw logits, but because softmax is invariant to constant shifts, a model's logit zero-point is arbitrary and left unconstrained by training. The authors demonstrate two measurable consequences: the penalty is effectively undefined across models (re-centering logits changes 58–96% of greedy tokens at a routine theta=1.3, while alternative penalties change none), and it severely corrupts structured output (dropping valid JSON schema conformance from 97% to 23% across 200 real-world schemas). Experiments spanning five models up to 7B parameters, two code models on HumanEval and JSONSchemaBench, and replication inside vLLM and llama.cpp confirm both effects. The authors show that applying the penalty to normalized log-probabilities instead of raw logits removes both problems, and note that HuggingFace already ships this operator (LogitNormalization) but applies it after the penalty and off by default.

← Back to the dashboard