Feb 2026· 2026 IEEE 23rd International Conference on Software Architecture Companion (ICSA-C)· pp. 341-345· 0 citations· 13 references
Computer Science
TL;DR
Carbon-Aware Governance Gates (CAGG), an architectural extension that embeds carbon budgets, energy provenance, and sustainability-aware validation orchestration into human-AI governance layers, is proposed.
Abstract
The rapid adoption of Generative AI (GenAI) in the software development life cycle (SDLC) increases computational demand, which can raise the carbon footprint of development activities. At the same time, organizations are increasingly embedding governance mechanisms into GenAI-assisted development to support trust, transparency, and accountability. However, these governance mechanisms introduce additional computational workloads, including repeated inference, regeneration cycles, and expanded validation pipelines, increasing energy use and the carbon footprint of GenAI-assisted development. This paper proposes Carbon-Aware Governance Gates (CAGG), an architectural extension that embeds carbon budgets, energy provenance, and sustainability-aware validation orchestration into human-AI governance layers. CAGG comprises three components: (i) an Energy and Carbon Provenance Ledger, (ii) a Carbon Budget Manager, and (iii) a Green Validation Orchestrator, operationalized through governance policies and reusable design patterns.
Amid the rapid growth of Artificial Intelligence (AI) adoption in business environments, concerns regarding energy consumption, carbon emissions, ethical accountability, and sustainable governance have become increasingly significant, creating the need for frameworks that align AI deployment with sustainability objectives and digital transformation strategies. This study aims to develop a green AI governance framework that supports sustainable digital business transformation by integrating environmental, governance, and technological considerations into AI implementation practices. Using a Systematic Literature Review (SLR), relevant studies published between 2020 and 2025 were identified, screened, and analyzed to synthesize key themes, governance dimensions, and sustainability principles associated with green AI adoption in organizational contexts. The findings reveal that effective green AI governance is shaped by four interconnected dimensions including environmental sustainability, ethical and regulatory compliance, operational efficiency, and stakeholder accountability. Based on these findings, a conceptual framework is proposed that links AI lifecycle management, governance mechanisms, sustainability metrics, and business transformation outcomes to promote responsible and resource efficient AI deployment. The study concludes that organizations can enhance the sustainability and long term value of digital transformation initiatives by adopting a structured green AI governance approach that balances innovation, transparency, and environmental responsibility. The proposed framework contributes to both academic literature and managerial practice by offering strategic guidance for implementing sustainable AI systems while supporting organizational competitiveness and broader Sustainable Development Goals (SDGs).
S. Syahyono, Rifqi Fahrudin, T. L. Anita et al.· ADI Journal on Recent Innova...· 0 citations
Artificial Intelligence (AI) is transforming the global business landscape by enabling organizations to improve efficiency, innovation, and sustainability performance. In recent years, businesses have increasingly integrated AI technologies into their operational and strategic activities to strengthen Environmental, Social, and Governance (ESG) practices and contribute toward achieving the United Nations Sustainable Development Goals (SDGs). This paper examines the role of AI in enabling sustainable business transformation and enhancing ESG performance.The study is conceptual in nature and is based on secondary data collected from journal articles, sustainability reports, industry publications, and global case examples. The research highlights how AI-driven technologies such as machine learning, predictive analytics, intelligent automation, and data analytics support organizations in resource optimization, carbon emission reduction, supply chain resilience, ethical governance, and stakeholder engagement. The findings indicate that AI can significantly improve sustainability performance by enabling data-driven decision-making, operational transparency, and efficient resource management. AI applications in smart energy systems, waste reduction, workforce management, and ESG reporting are helping organizations align business goals with sustainability objectives. Companies such as Microsoft, Tesla, and Unilever demonstrate how AI can be strategically integrated into sustainable business models to create long-term economic, environmental, and social value.However, the study also identifies key challenges including data privacy concerns, algorithmic bias, ethical risks, implementation costs, and workforce displacement. Therefore, organizations must adopt responsible AI governance frameworks and ensure transparency, accountability, and inclusiveness in AI adoption.
S. R, Neetha Veronica A, Arpita Sastri et al.· International journal of com...· 0 citations
The transition toward decentralized renewable energy systems creates monitoring problems that current digital infrastructures do not solve: sustainability claims are produced by the same actors they evaluate, environmental evidence is reported periodically rather than observed continuously, and the communities most affected by deployment cannot inspect the data used to represent their territories. Existing integrated platforms combine subsets of the blockchain, Internet of Things (IoT) sensing and life cycle assessment (LCA) at the data layer, but they do not organize that integration through an explicit governance structure. This paper contributes a cybernetic governance framework in which the Viable System Model (VSM) supplies the organizing structure of a blockchain–IoT–LCA monitoring architecture, so that sensing, distributed trust, strategic intelligence and participatory governance are recursively coupled rather than sequentially chained. The framework was developed and evaluated under the Design Science Research paradigm, and instantiated in the IMPACT Energy.CO platform across two technology routes, wind and solar, in La Guajira, Cesar, Atlántico and Magdalena, Colombia. Evaluation against six pre-declared criteria reports 45 executed test cases with a 100% pass rate, 90% unit and 87% integration code coverage, load tests up to 5000 concurrent users with zero errors and sub-second mean response, an operating hash-chained provenance layer issuing verifiable LCA certificates, 14 participatory validation workshops, 199 users trained and 166 technicians certified. We use traceability in a deliberately narrow sense throughout: the property whereby a committed record can be linked to the ingested data series, model version and computation that produced it, and its integrity and ordering checked by a party that does not trust the producer. It is provenance and integrity traceability from the point of ingestion onward, and it is not metrological traceability: the architecture cannot verify that an original sensor measurement corresponds to the physical quantity it purports to represent. We accordingly make explicit what the architecture does not guarantee: a ledger protects records after commitment but cannot certify measurement at the point of capture, and we present a threat model, a set of implemented controls and the residual risk that remains. This study contributes an architecture, a reproducible development and evaluation method, and a calibrated account of what verifiable environmental monitoring can and cannot deliver in contested Global-South territories.
J. Taborda, Cesar Enrique Polo Castro, Alexander Armando Bustamante et al.· Future Internet· 0 citations
As a result of this rapid pace of digital enterprise transformation (powered by AI, distributed cloud, and customer-facing digital experience platforms), there is now a serious corporate sustainability paradox: energy use and Scope 3 carbon emissions. There are technical solutions for instance local Green AI, carbon-conscious resource management and automatic emission monitoring, but very fragmented, limited to departmental pockets of IT where there is no macro level business blueprint. That leaves business leaders constantly caught in the dilemma of introducing high level concepts of Environmental, Social and Governance (ESG) into digital systems.
To fill this execution gap, this study suggests a complete and socio-technical framework for Digital Experience Carbon-Aware Enterprise Architecture, called Green by Design. The framework is cantered on Design Science Research (DSR) paradigm and systematic themed analysis of the existing practices in the field of management information systems and combines the concepts of corporate governance, transparency of data, optimization of assets and perpetual life cycle verification into an actionable corporate strategy. The method is based on organizational paradigms (Industry 5.0 philosophy, Social Technological Systems Theory and Value Sensitive Design (VSD)), and on operational tools (Digital Twins and Retrieval-Augmented Generation (RAG) enabled carbon accounting).
The results show that the multi-layered approach to embedding carbon accounting into enterprise architecture can cut data infrastructure carbon footprints by up to 56% and optimise capital expenditure and cloud asset allocation. Moreover, the use of intelligent carbon systems into data validated systems further improves the accuracy of corporate reporting and environmental reporting beyond 90%, mitigating the risk of legal compliance and helping to avoid “greenwashing” of brands. This is a blueprint for the C-suite, Chief Information Officers (CIOs) and sustainability directors to help their companies move forward on the technological side of the business to support sustainability and long-term value for their stakeholders.
Nik Abdullah Bin Rozali, Amli Omar Bin Ismail, Sherry Ameera Binti Mustaffa @ Sulaiman et al.· International journal of res...· 0 citations
Artificial intelligence (AI)-enabled systems, such as large language models (LLMs) and federated learning, are fundamentally transforming organisational workflows into distributed, model-centric ecosystems. While these advancements drive innovation, they simultaneously heighten information governance (IG) challenges regarding data privacy, algorithmic accountability, and systemic transparency. Although various ethical frameworks have been proposed to address these concerns, their practical operationalisation remains fragmented across different sectors. This study addresses this gap through a PRISMA 2020 guided systematic review of 78 peer-reviewed studies published between 2020 and 2024, aiming to synthesise a cohesive architecture for modern AI governance. The review identifies five core IG functions essential for maintaining integrity: accountability, data quality, privacy-by-design, compliance, and risk management. These functions are not merely theoretical; they are enabled by six distinct technological clusters, including privacy-preserving federated learning, machine-learning operations (MLOps), and AI–blockchain traceability. These tools allow organisations to move beyond manual oversight towards automated, scalable governance. Furthermore, the research highlights that human governance must be distributed across ethical oversight and technical validation. This necessitates a multidimensional competency architecture that spans legal awareness, analytical proficiency, and socio-technical skills. Ultimately, IG is evolving from a reactive, post hoc compliance exercise into a proactive, lifecycle-oriented socio-technical capability. By integrating advanced technological clusters with human-centred agency, organisations can bridge the gap between abstract ethical principles and the practical, distributed demands of modern AI governance. This holistic shift ensures the creation of robust, transparent, and secure information ecosystems that are resilient to the complexities of the digital age.
A. Faza, Ilyana Agri Lestari· Applied Cybersecurity &...· 0 citations
Although software systems increasingly shape energy consumption, economic output, and societal welfare, most development methods still prioritise schedule, cost, and functionality. This paper presents an approach to software development that prioritises sustainability by embedding environmental, economic, and social objectives into the SDLC from the very beginning. The framework represents sustainability as a set of quantifiable variables that are combined into a Global Sustainability Index (GSI). These metrics include operational energy and carbon footprint, total cost of ownership (TCO), maintainability index, defect density, and a normalised Social Impact Score (SIS). An empirical measurement architecture gathers runtime and process data for continuous improvement, while phase-level “sustainability budgets” direct trade-offs across requirements, design, implementation, testing, and operation. The framework is evaluated in a repeated-measures industrial study of six production software systems (758 KLOC in total, 58 engineers, six application domains), in which every system is observed over four counterbalanced release cycles governed respectively by the proposed framework and by three established approaches: GREENSOFT, GreenSDLC, and the Sustainability Quality Model (SQM). The proposed framework attains the highest GSI (0.86 ± 0.03), a statistically significant improvement of 10–19% over the competing frameworks (paired t-tests, all Holm-adjusted p < 0.002, Cohen’s dz > 2.5). Relative to current models, energy usage and carbon emissions are cut by 10–20%, and they are decreased by approximately 30% when compared to a no-framework baseline. Normalised maintainability and defect density both improve over a five-year timeframe, and total cost of ownership drops 4–9%. Consistently higher levels of social impact and stakeholder satisfaction are observed, particularly for user groups who are marginalised. Ninety-five percent confidence intervals and effect sizes are reported for every headline comparison, and the principal limitations of the framework are stated explicitly together with mitigation strategies. These results show that all three dimensions can be improved with explicit quantitative sustainability integration without a rise in long-term costs.
M. K., P. Pareek· International Journal of Adv...· 0 citations
Related blog posts
Microsoft Research Blog· microsoft.comJul 30, 2026
Computer-use AI agents struggle with multi-step workflows like email and customer support. Echoverse trains agents in realistic environments rather than simply providing more training tasks, helping them improve as the tasks, tests, and environments evolve. The post Echoverse: Deep, evolving environments for computer-use agents appeared first on Microsoft Research.
AI is making software generation faster, but speed does not remove the need for expertise. As more work is delegated to AI, tacit knowledge may become one of the most important human advantages in software engineering. The post Beyond Prompt Engineering: The Role of Tacit Knowledge in Software Engineering appeared first on GPT-Lab.
MIT News · Artificial Intelligence· news.mit.eduJun 30, 2026
Computer scientist Phillip Isola cuts through the hype to explain how AI agents work and what the future might hold for this rapidly advancing technology.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
