Frontiers in Artificial Intelligence最新文献

{"title":"Explainable AI for phishing URL detection: a Bayesian-optimized stacking ensemble framework with SHAP-guided feature learning.","authors":"Hafiz Aziz Khan, Sonia Akter, Abdur Rahman Lindon, Taslima Akter, Iftekhar Rasul, Mamunur Rahman, Nasrin Akter Tohfa, Iftekhar Hossain","doi":"10.3389/frai.2026.1854934","DOIUrl":"10.3389/frai.2026.1854934","url":null,"abstract":"<p><strong>Introduction: </strong>Phishing remains one of the most persistent and financially damaging threats facing modern organizations, with over 4.7 million incidents recorded in 2023 alone. Existing AI-based phishing detection frameworks are constrained by limited benchmarking scope, absent model interpretability, and insufficient statistical validation - three limitations that collectively restrict operational utility in real-world security environments.</p><p><strong>Methods: </strong>We present an explainable, end-to-end machine learning pipeline evaluated on a large public benchmark of 247,950 URLs described by 41 structural and lexical features. The pipeline integrates SHAP-driven feature selection (reducing 41 to 24 features via a 95% cumulative-signal rule), a systematic benchmark of 12 classifiers spanning seven algorithmic families, Bayesian hyperparameter optimization via Optuna TPE sampling (40 trials each for XGBoost and CatBoost), and a heterogeneous stacking ensemble combining Optuna-tuned XGBoost, CatBoost, Extra Trees, and Random Forest under a logistic-regression meta-learner. A four-layer statistical validation protocol - comprising a Friedman omnibus test, Wilcoxon signed-rank tests, paired t-tests, and Cohen's d effect sizes - was applied to five-fold cross-validation accuracy distributions to assess directional consistency, with the limited inferential resolution of five folds explicitly acknowledged.</p><p><strong>Results: </strong>SHAP-driven selection reduced the feature space by 41.5% while retaining 95% of predictive signal. The stacking ensemble achieved 96.75% accuracy, 96.74% F1-score, and AUC of 0.9947, attaining the lowest Brier score among all 13 models (0.0246), indicating superior probability calibration. The Friedman omnibus test confirmed significant performance differences across models (χ²F = 59.84, <i>p</i> < 0.0001), and all 12 Wilcoxon pairwise comparisons yielded the minimum attainable p-value (<i>p</i> = 0.0313), confirming the ensemble never lost a cross-validation fold against any baseline. Post-hoc SHAP analysis identified subdomain structure, URL length, and URL entropy as the dominant phishing indicators at both ensemble and base-learner levels.</p><p><strong>Discussion: </strong>The co-leaders - the stacking ensemble and Extra Trees - demonstrate that rigorous, interpretable AI pipelines can advance phishing detection accuracy and transparency simultaneously. The framework's calibrated risk scores, threshold flexibility, and multi-level SHAP explainability support analyst-facing decision-making in security operations, while its leakage-free.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1854934"},"PeriodicalIF":6.7,"publicationDate":"2026-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13672667/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216954","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Intelligent defense at the edge: a comprehensive survey of federated learning, TinyML and explainable AI for intrusion detection in IoT and IIoT ecosystems.","authors":"S Sarath Kumar, M Asha Jerlin","doi":"10.3389/frai.2026.1900815","DOIUrl":"10.3389/frai.2026.1900815","url":null,"abstract":"<p><p>The proliferation of Internet of Things (IoT) and Industrial Internet of Things (IIoT) technologies has fundamentally transformed contemporary computing infrastructures by interconnecting large heterogeneous devices, sensors, embedded systems, and cyber-physical platforms. These ecosystems support diverse applications including smart cities, intelligent transportation, industrial automation, medical surveillance, precision agriculture and home automation. Despite their operational advantages, the vast connectivity and inherent resource constraints of IoT devices significantly expand the cyberattack surface. Limited processing power, limited memory and restricted energy budgets render conventional cybersecurity solutions infeasible for IoT environments. Consequently, IoT systems have become prime targets for advanced threats, including malware, botnets, Distributed Denial of Service (DDoS) attacks, ransomware, insider attacks and data manipulation. Traditional Intrusion Detection Systems (IDS), designed for enterprise networks, rely on centralized data collection and computationally intensive machine learning models that are unsuitable for IoT deployments due to scalability limitations, communication overhead, latency constraints and privacy concerns. Recent research has therefore shifted toward lightweight, intelligent and distributed cybersecurity frameworks capable of providing robust protection under severe resource limitations. This survey comprehensively reviews recent advances in AI-based intrusion detection and malware analysis for IoT ecosystems, covering lightweight machine learning models (ML), deep learning (DL) architectures for edge deployment, federated learning (FL) based privacy-preserving detection, TinyML on-device intelligence, blockchain-based trust management, Explainable AI (XAI) driven transparent monitoring and adaptive cyber defense. A unified conceptual framework is proposed integrating lightweight detection, distributed collaborative learning, trust-based orchestration and adaptive defense.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1900815"},"PeriodicalIF":6.7,"publicationDate":"2026-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13673141/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Context-aware KazRoBERTa with transformer-based candidate re-ranking for Kazakh spelling correction.","authors":"Zhanar Azhibekova, Daniyar Sultan, Laura Baitenova","doi":"10.3389/frai.2026.1925050","DOIUrl":"10.3389/frai.2026.1925050","url":null,"abstract":"<p><p>Automatic spelling correction for Kazakh is difficult because productive morphology, vowel harmony, and Cyrillic keyboard substitutions generate many orthographically plausible alternatives. This study proposes Context-Aware KazRoBERTa, a three-stage framework that combines token-level error detection, linguistically constrained candidate generation, and masked-language-model candidate re-ranking with a shared Transformer encoder. Candidate generation integrates inverse ka2ru mapping, vowel-harmony restoration, Optimal String Alignment distance, and lexicon filtering; contextual probability, orthographic similarity, and lexical frequency are then combined for selection. Evaluation was performed on a synthetic corpus of 2,615,771 parallel sentence pairs derived from 922,798 clean sentences. On the held-out synthetic test set, the framework achieved 97.7% detection F1-score and 98.8% detection accuracy, while sentence correction accuracy reached 85.8%, word correction accuracy reached 95.6%, and character error rate decreased to 1.86%. The 95% Wilson interval for sentence correction accuracy was 85.36-86.23%. Relative to the reported two-stage baseline, the 5.6-point sentence-accuracy margin implies a conservative continuity-corrected McNemar lower bound of χ² > = 230.3 (<i>p</i> < 0.001) across all paired contingency tables compatible with the reported rounded margins. Ablation results attribute the largest gain to contextual re-ranking, particularly for deletion and substitution errors. The evidence nevertheless remains limited to one fixed-seed synthetic evaluation: no natural-error test corpus, direct ByT5/mT5/NLLB or sequence-tagging baseline, paired prediction archive, repeated-run variance, or measured latency benchmark was supplied. The web interface therefore demonstrates system integration rather than external accuracy. The scientific contribution is the Kazakh-specific, parameter-sharing hybrid design; claims of real-world generalization or state-of-the-art superiority require external benchmarking and fully archived configurations.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1925050"},"PeriodicalIF":6.7,"publicationDate":"2026-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13672556/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216746","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Artificial intelligence, explainable deep learning, and generative models for osteoporosis and bone disease prediction: a systematic review and future research roadmap.","authors":"M Raja, Avulapalli Jayaram Reddy","doi":"10.3389/frai.2026.1930437","DOIUrl":"10.3389/frai.2026.1930437","url":null,"abstract":"<p><p>Osteoporosis is a major global health burden, with considerable morbidity, mortality, and healthcare costs. Artificial intelligence (AI) has opened new avenues for osteoporosis screening, bone mineral density quantification, fracture-risk prediction, and clinical decision support, but progress remains fragmented across machine learning, deep learning, explainable AI (XAI), generative models, and multimodal learning. This systematic review brings these strands together and identifies priority research needs for AI-based osteoporosis prediction. Following the PRISMA 2020 statement, studies were identified from Scopus and Google Scholar, published between January 2020 and the final search date of 30 June 2026, with two reviewers screening independently and resolving disagreements by consensus. Design-related risk-of-bias indicators, such as retrospective vs. prospective design, validation strategy, and reference standard, were recorded for every study; a full PROBAST, QUADAS-2, and Newcastle-Ottawa Scale appraisal is planned but not yet complete. Fifty-eight studies met the inclusion criteria and were synthesized narratively and descriptively, without formal meta-analysis, given substantial heterogeneity. Findings spanned five themes: classical machine learning, deep learning for imaging, explainable AI, generative models, and multimodal learning. Deep learning achieved strong predictive accuracy (AUC 0.82-0.96), XAI improved clinical interpretability, generative models helped offset data scarcity, and multimodal approaches boosted performance by combining imaging and clinical data. Key gaps remain, including limited external validation, heavy reliance on retrospective data, poor generalisability, and limited regulatory readiness. Building on these findings, this review proposes MXGNet, a multimodal, explainable, generative framework incorporating privacy-preserving federated learning, intended to guide development of robust, interpretable, clinically deployable AI tools for osteoporosis prediction.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1930437"},"PeriodicalIF":6.7,"publicationDate":"2026-10-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13672560/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Self-evaluative reliability: a missing safety property in adaptive clinical AI governance.","authors":"Rasit Dinc, Nurittin Ardic","doi":"10.3389/frai.2026.1973339","DOIUrl":"10.3389/frai.2026.1973339","url":null,"abstract":"","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1973339"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13668799/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149217017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing vehicle routing problem through a qubit-efficient Symmetric Arc Reduction using Quantum Approximate Optimization Algorithm.","authors":"Jayaraja A, Senthil Kumaran U","doi":"10.3389/frai.2026.1900887","DOIUrl":"10.3389/frai.2026.1900887","url":null,"abstract":"<p><strong>Introduction: </strong>Finding the most efficient routes for a fleet of delivery vehicles remains a critical and costly challenge for nearly any logistics business. This core idea is known as the Vehicle Routing Problem (VRP). The primary goal is to minimize operational costs, such as the total travel distance or the number of trucks deployed, while ensuring all the customers are served. Because VRP is NP-hard, finding a perfect solution is often impossible, forcing industries to rely on classical approximate algorithms. However, quantum variational algorithms offer a new paradigm by iteratively optimizing parameters to navigate the solution space, presenting a new and promising avenue for finding better and faster solutions.</p><p><strong>Methods: </strong>In this paper, we introduce Symmetric Arc Reduction in Quantum Approximate Optimization Algorithm (SAR-QAOA), which systematically eliminates redundant symmetric routing variables in the binary integer programming formulation prior to the Quadratic Unconstrained Binary Optimization (QUBO) formulation. The proposed SAR-QAOA encodes edge-based VRP while reducing the qubit requirement from <i>n</i>(<i>n</i>-1) to <i>n</i>(<i>n</i>-1)/2, thereby improving the resource efficiency of the QUBO formulation by prioritizing regions of the solution space that are more likely to contain high-quality solutions; SAR enhances both search accuracy and computational efficiency.</p><p><strong>Results: </strong>Experimental results on noiseless quantum simulators validate the feasibility of the proposed formulation and its integration with the QAOA framework, providing a promising foundation for future investigations on larger routing problems and evolving quantum hardware.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1900887"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13669089/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216950","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An AI-driven hybrid random forest-LSTM framework for IoT-based groundwater level prediction and sustainable water management in Chennai, India.","authors":"S Boomika, R Sitharthan, R Tanukasree, R Senthil Kumar","doi":"10.3389/frai.2026.1958933","DOIUrl":"10.3389/frai.2026.1958933","url":null,"abstract":"<p><strong>Introduction: </strong>Groundwater resources in Chennai, Tamil Nadu, India are under increased stress due to rapid urbanization, erratic monsoon patterns and largely unregulated groundwater extraction. Addressing this challenge requires not only continuous monitoring but also accurate forecasting of groundwater fluctuations.</p><p><strong>Methods: </strong>This study presents a data driven framework that combines Internet of Things (IoT) enabled sensing with Machine Learning model. Real-time measurements such as groundwater depth, rainfall, temperature, and soil moisture are combined with historical hydro-meteorological data to build a reliable prediction system. A hybrid modelling approach is adopted in this study, where Random Forest is used to identify key influencing factors and capture non-linear relationships, while Long Short-Term Memory (LSTM) networks are employed to model temporal variations.</p><p><strong>Results: </strong>The combined RF-LSTM model surpasses traditional techniques, delivering an RMSE of 0.38 m, an MAE of 0.29 m, and R² value of 0.96. Relative to a standalone LSTM model, the hybrid approach cuts the RMSE by approximately 22%, and it also outperforms Random Forest, Support Vector Regression, and Linear Regression individually.</p><p><strong>Discussion: </strong>The model successfully tracks seasonal shifts in groundwater capturing groundwater recharge during the Northeast monsoon and depletion during the summer season. Through timely and dependable forecasts, this system aids more effective groundwater management, strengthens rainwater harvesting efforts and supports sustainable urban water practices going forward.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1958933"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13669168/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Endophytic actinomycetes as a reservoir of novel chemotherapeutic leads against multidrug-resistant cancer.","authors":"Dhanvee Ganatra, Mahendrapal Singh Rajput, Gaurav Sanghvi","doi":"10.3389/frai.2026.1974583","DOIUrl":"10.3389/frai.2026.1974583","url":null,"abstract":"<p><p>Multidrug resistance (MDR) has emerged as a major barrier to effective cancer control, undermining the efficacy of chemotherapy, targeted agents and, increasingly, immunotherapies. This review synthesizes current knowledge on the molecular and cellular bases of MDR in cancer, including ABC transporter-mediated drug efflux, enhanced DNA damage responses, survival-signaling rewiring, tumor-microenvironmental protection and cancer stem cell-driven plasticity. Particular emphasis is placed on endophytic actinomycetes and their metabolites as an underexploited reservoir of MDR modulating chemotypes with potent cytotoxic, resistance-reversal and microenvironment-modulating activities. By integrating recent mechanistic and pharmacological data, the review first delineates the crisis of drug-resistant cancer, then catalogs actinomycete-derived metabolites with documented efficacy against resistant cell lines, including those that inhibit efflux pumps, restore pro-apoptotic signaling or interfere with angiogenesis, epithelial-mesenchymal transition and stemness traits. Finally, translational challenges, such as suboptimal pharmacokinetics, off-target toxicity, herb-drug interactions and the paucity of biomarker-driven clinical trials, are critically examined alongside emerging strategies involving semi-synthetic optimization, nanocarrier delivery and rational combination regimens. A forward-looking discovery pipeline is proposed that couple's genome mining of endophytic actinomycetes with systematic MDR-mechanism profiling and medicinal chemistry, aiming to prioritize the most promising metabolites for preclinical and clinical development as next-generation MDR-targeted therapeutics.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1974583"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13668826/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216909","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"UniPET: a unified approach for whole-body CT to PET translation.","authors":"Francesco Di Feola, Valerio Guarrasi, Mikael Johansson, Paolo Soda","doi":"10.3389/frai.2026.1919564","DOIUrl":"10.3389/frai.2026.1919564","url":null,"abstract":"<p><p>Positron emission tomography (PET) provides critical metabolic information for oncological imaging, yet its use is constrained by radiation exposure, cost, and limited availability. Synthesizing PET-like images from computed tomography (CT) has been proposed as a way to approximate metabolic information; however, existing approaches either fail to capture region-specific variability or rely on multiple organ-specific models that do not scale to whole-body imaging. In this work, we introduce UniPET, a unified approach for whole-body CT-to-PET translation based on curriculum learning. By progressively incorporating anatomical regions of increasing complexity, UniPET enables a single network to learn coherent morpho-metabolic relationships across different regions. We evaluate UniPET on a public dataset of 900 patients and an external cohort of 579 patients. The model achieves image quality and metabolic consistency comparable to region-specific approaches, while improving over conventional whole-body models and maintaining stable performance under distribution shifts. By capturing region-specific variability within a single model, UniPET provides a promising approach for augmenting CT-based workflows with synthetic metabolic information, with potential applications in screening triage, in low-resource environments where PET is unavailable, and as an additional input for multimodal analysis pipelines. The code is available at: https://github.com/arco-group/UniPET.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1919564"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13668879/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149216955","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Large-scale conversational data auditability via cryptographic commitments and hash-chained audit-ledger verification.","authors":"Iván Ortiz-Gárces, Pablo Palacios, Washington Loza-Herrera, William Villegas-Ch","doi":"10.3389/frai.2026.1929127","DOIUrl":"10.3389/frai.2026.1929127","url":null,"abstract":"<p><p>The auditability of large language model-based systems requires verifiable evidence without exposing the sensitive content of interactions. This article proposes a hash-chained audit-ledger framework to transform public LLM conversations into canonical off-chain units, generate cryptographic commitments using SHA-256 and HMAC-SHA-256, build a lightweight ledger with Merkle roots, and verify integrity against induced tampering. The study integrates three heterogeneous sources, LMSYS-Chat-1M, WildChat, and Chatbot Arena, and produces 3,971,887 auditable units. In the full-corpus HMAC-SHA-256 experiment, the 11,899.55 MiB normalized repository was represented by 1,473.44 MiB of hash-only audit records, 2.03 MiB of ledger block headers, and 380.52 MiB for the operational ledger plus membership structures required for record-level verification. Five induced tampering categories, Merkle and decision-hash integrity, hash-chain linkage integrity, block ordering integrity, membership completeness, and dataset-membership consistency, were detected without false acceptance; a separate clean-ledger baseline validated the untampered ledger with zero error signals. Load experiments showed near-linear growth of mean total pipeline time with respect to the number of audit records. Operational ledger-stage throughput increased across the intermediate workloads but declined at the largest evaluated load, while full record-set verification throughput remained comparatively stable up to 250,000 interactions before decreasing in the 500,000-interaction scenario. This contribution demonstrates that the traceability of decisions generated by LLM can be formulated as a measurable, private, and cryptographically verifiable experimental chain. The framework provides a documented numerical-audit basis for independent consistency assessment, regulatory compliance, and forensic analysis when ledger states are evaluated against preserved artifacts, trusted checkpoints, or externally anchored commitments.</p>","PeriodicalId":33315,"journal":{"name":"Frontiers in Artificial Intelligence","volume":"9 ","pages":"1929127"},"PeriodicalIF":6.7,"publicationDate":"2026-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13668802/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149217033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}