Jul 2026· International Journal of Engineering Science and Information Technology· Vol 6, pp. 153-160· 0 citations· 33 references
TL;DR
Two complementary governance frameworks that establish a structured model for human–AI collaboration in enterprise software testing by improving testing efficiency, auditability, transparency, regulatory compliance, and organizational confidence in AI-supported continuous delivery practices without compromising human oversight or decision accountability are proposed.
Abstract
Enterprise software testing organizations must balance the scalability required by continuous delivery with the contextual judgment necessary for effective quality assurance. Although automated testing enables rapid execution and extensive regression coverage, it lacks the domain expertise, business context, risk awareness, and ethical accountability required for critical release decisions. This paper proposes two complementary governance frameworks that establish a structured model for human–AI collaboration in enterprise software testing. The Human–AI Responsibility Allocation (HARA) Model defines the optimal distribution of testing activities based on comparative strengths, assigning repetitive and computationally intensive tasks—including regression testing, pattern recognition, anomaly detection, and test execution—to artificial intelligence, while reserving strategic responsibilities such as test planning, defect prioritization, release readiness assessment, governance, and compliance oversight for human experts. To operationalize this allocation, the AI Confidence-Based Escalation Framework (ACEF) introduces a three-tier escalation mechanism that dynamically determines when AI-generated testing outcomes require human review according to confidence scores, business criticality, and organizational risk tolerance. The framework further incorporates measurable governance indicators, including escalation rate, false-positive rate, human override frequency, model drift, and decision traceability, enabling continuous monitoring of AI performance and accountability. The proposed frameworks are evaluated conceptually across regulated enterprise environments, including insurance, financial services, and healthcare, where software quality directly affects regulatory compliance, operational resilience, and customer trust. The analysis demonstrates that clearly defined accountability boundaries enable organizations to achieve the speed and scalability of AI-assisted testing while preserving human judgment for high-risk decisions. The proposed governance architecture provides a practical foundation for responsible AI adoption in software quality assurance by improving testing efficiency, auditability, transparency, regulatory compliance, and organizational confidence in AI-supported continuous delivery practices without compromising human oversight or decision accountability
Agile software development emphasizes rapid iteration, continuous integration, frequent releases, and incremental delivery, making regression testing a central software quality challenge. Conventional regression testing approaches often depend on manually selected test suites, static prioritization rules, and repeated execution of tests that provide limited incremental fault-detection value. This paper develops a conceptual intelligent regression testing framework that applies artificial intelligence (AI) techniques to test selection, prioritization, execution, failure classification, and continuous learning within Agile development pipelines. The methodological foundation combines supervised learning, representation learning, historical test-result analysis, change-impact assessment, and feedback-driven optimization. Because the supplied literature primarily concerns AI-based detection and classification in biomedical signal-processing applications rather than software testing, the paper explicitly treats these studies as methodological evidence for transferable AI patterns rather than direct empirical evidence for regression testing. The framework consequently emphasizes feature extraction, automated classification, adaptive prediction, and real-time decision support. A conceptual evaluation indicates that AI-assisted regression testing can improve the alignment between code changes and test execution priorities, reduce redundant execution, and create feedback loops capable of adapting to changing Agile projects. However, model drift, insufficient historical data, explainability, false prioritization, and integration complexity remain significant constraints. The analysis positions intelligent regression testing as an adaptive decision-support layer rather than a complete replacement for conventional testing practices.
Denis Kazlauskas· The American Journal of Engi...· 0 citations
Empirical evidence is provided that generative AI can effectively support security requirements engineering when embedded within human-centered workflows and organizational governance structures, offering practical insights for adoption in regulated software development contexts.
F. Martins, Elaine Venson· SIGSOFT FSE Companion· 0 citations
Generative Artificial Intelligence systems—particularly those built on Large Language Models (LLMs)—have become central to modern enterprise computing, yet they carry with them a class of vulnerabilities that traditional cybersecurity models were never designed to address. Decoder-only transformer architectures process system instructions and untrusted user inputs as a single undifferentiated sequence of tokens, which makes them susceptible to direct and indirect prompt injections, jailbreaking attacks, and retrieval poisoning. Existing governance frameworks—including the NIST AI Risk Management Framework, the EU AI Act, and ISO/IEC 42001—offer valuable compliance guidance at the organizational level, but they stop short of providing the kind of execution-level security blueprints that engineering teams actually need.
This paper introduces the Prompt-Level AI Risk Governance Framework (PLAIRGF), a four-phase architectural model covering Prevention, Detection, Response, and Continuous Improvement. Rather than depending on simulated metrics to evaluate the framework, we validate PLAIRGF through mathematical formalization of key risk indicators—including Attack Success Rate (ASR), Risk Severity Score (R), the Governance Readiness Index (GRI), and a newly proposed Human-in-the-Loop Alignment Coefficient (η)—combined with rigorous defensive capability mapping and step-by-step operational trace walkthroughs of representative adversarial attack scenarios. The paper concludes with a structured alignment between PLAIRGF controls and international compliance standards, offering organizations a practical, audit-ready foundation for deploying LLM-based systems securely.
Keywords: Generative AI Security, Large Language Models, Prompt Injection, AI Governance, Retrieval-Augmented Generation, Human-in-the-Loop, PLAIRGF
Dr. Abdul Majid Farooqi Dr. Abdul Majid Farooqi, Ziya Anjum Ziya Anjum· International Scientific Jou...· 0 citations
The increasing complexity, scale, and release frequency of contemporary software systems have exposed limitations in conventional testing practices, particularly in exhaustive test execution, regression validation, and early defect identification. This research proposes an intelligent framework that integrates artificial intelligence (AI)-based test automation with software defect prediction to establish a proactive quality-engineering process. The proposed framework combines requirement and code analysis, automated test generation, execution prioritization, defect-risk estimation, feedback-driven model refinement, and quality reporting within a unified architecture. The methodological foundation is a conceptual synthesis of AI-driven test automation principles, with particular emphasis on automation, intelligent prioritization, and predictive quality assurance as discussed by Ramamurthy (2023). The supplied reference corpus also demonstrates how intelligent sensing, pattern recognition, resource optimization, and data-driven classification can conceptually inform automated quality-monitoring architectures, although most of these studies originate outside software testing. The proposed model therefore treats cross-domain evidence as methodological inspiration rather than direct empirical validation. The analysis indicates that combining predictive defect-risk scores with automated test selection can potentially reduce redundant testing, concentrate computational resources on high-risk software components, and improve feedback speed. However, model reliability depends on historical defect data, feature quality, distributional stability, explainability, and integration with existing development pipelines. The framework contributes a structured basis for AI-assisted software quality engineering while identifying empirical validation, benchmark datasets, and explainable prediction as priorities for future research.
Haruto Tanaka, Yuki Nakamura· Frontiers in Emerging Multid...· 0 citations
The deployment of artificial intelligence systems in enterprise risk management presents a novel governance challenge: determining which AI-generated risk assessments carry sufficient confidence for automated action and which warrant human review. This is the confidence-conditioned escalation problem, where uncertainty quantification governs the boundary between automated AI and human intervention. Current enterprise governance practices rely on categorical automation rules that conflate risk category classification with model confidence, creating a critical epistemic gap. This paper formalizes the Confidence-Aware Escalation (CAE) framework, which integrates quantified uncertainty into escalation decisions via conformal prediction theory. Conformal prediction provides distribution-free, finite-sample coverage guarantees, enabling governance policies to be specified in terms of verifiable error rate bounds rather than heuristic confidence thresholds. The CAE framework classifies AI outputs into three automation tiers by jointly evaluating prediction-set cardinality and risk category. A three-layer governance architecture comprising inference production, human oversight, and regulatory compliance is proposed, supported by structured stakeholder roles: Risk Owners, Model Stewards, and Executives. Adaptive feedback learning pipelines maintain coverage guarantees under concept drift without reward-hacking incentives. A policy-driven fairness monitoring protocol resolves mathematical incompatibility among fairness criteria through organizational policy specification rather than technical compromise. Empirical validation on a publicly available financial risk benchmark dataset confirms that the framework achieves high automation rates while maintaining near-theoretical coverage guarantees and enabling transparent governance of false escalation risk. The framework is aligned with EU AI Act high-risk classification requirements and US Federal Reserve SR 11-7 model risk management guidance.
Janardhana Naidu Kola· International journal of com...· 0 citations
The increasing availability of Artificial Intelligence (AI) tools has generated significant interest within the project management community; however, structured guidance tailored specifically to Project Managers remains limited. This paper proposes a standardized AI-enabled prompt architecture aligned with the PMBOK® 8E performance domains and processes. The framework consists of a Master Prompt, Task-Level Executable Prompts, and Refine Prompts designed to create bounded, context-aware interactions between Project Managers and AI systems. The proposed architecture embeds PMBOK-aligned terminology and follows the Inputs–Tools–Outputs (ITTO) logic. The AI system functions as an analytical and generative tool within this structure, processing structured and unstructured inputs—including expert judgment—and producing standardized outputs for managerial review and refinement. The architecture is designed to be generalized and extensible across all forty project management processes. This study does not present empirical performance metrics; rather, it introduces a structured conceptual framework intended for practical application and future validation. Practitioners are encouraged to implement the architecture in real project environments to evaluate measurable improvements and contribute to further academic development in AI-enabled project management.
Vittal Anantamula, Rajendra Harsh· 2026 6th International Confe...· 0 citations