ESG Part 3 – Enterprise Architecture Engagement in the Responsible Use of AI in the Evolving ESG Landscape

(Editor’s Note: What follows is Part 3 of a 3-part article, July 1. The abstract for the 3-part series appeared here. The author, Lisa A. Pratico, was profiled here in the spring of 2023.)

By Lisa A. Pratico

In Part 3 of this research paper, I look at why Enterprise Architecture must be engaged to ensure AI is being designed and deployed responsibly within the business to ensure maximum value with the proper techniques and deployment as well as assuring the organization is meeting its sustainability mandates.

In Part 2 I examined the growing push to quickly harness AI to speed up and scale efforts and find solutions to common challenges across businesses, in education and in society in general. Engaging an architecture competency with your lines of business, as well as at the corporate level, at the start will ensure that business cases answer critical questions on the value and ROI of the AI solution while also aligning to standard technology frameworks and new sustainability measures for regulatory reporting. Ironically, the involvement of Architecture and the skills brought to bear are no different than the structures and organizational models defined in existing “modern day” enterprise architecture groups as well as those “morphed” over time to ensure the appropriate usage and deployment of RPA or migration to modern-day cloud.

What is different is the frameworks being defined for ensuring sustainability which both architects and engineers need to be trained on and apply to their roles as AI takes hold across all industries. And it is taking hold! So, terms like Green EA1and Green Coding2standards are starting to pop up in IT as both architecture and engineering roles engage with new skills and assessments, patterns and principles. Green Coding and deployment standards are being integrated into modern day solutions. Sustainable IT goals must be evaluated alongside an organization’s business goals of value, maximizing ROI and having the potential to cause significant positive disruption. This means the architecture and engineering practice must take an in-depth look at how the business wants to solve problems with AI and assess the technology’s impact on environmental and ethical concerns around transparency and fairness.

As per, incorporating ESG (Environmental, Social, and Governance) sustainability principles into enterprise architecture is not just about doing what’s right for the planet and society; it’s about future-proofing businesses and fostering long-term digital resilience. By aligning enterprise technology strategies and design with ESG principles, organizations can drive positive impact while unlocking new opportunities for innovation, efficiency and competitive advantage. See “Sustainability Principles for Enterprise Architecture” by We have started to see this impact not just in how AI has gone from applied artificial intelligence in the 1990s with Deep Blue by IBM, but quickly advanced into NLP and ML. The next wave came with Deep Learning through artificial neural networks, which then was applied to Generative AI, which brings us to what everyone is talking about today.

The Need for Green EA and its linkage to other competencies to make Sustainable IT a Reality

The new composition of IT leadership to make Sustainable IT a reality must take hold to ensure sustainability strategies have equal airtime with executives and are considered with weight along with other IT strategies tied to digital transformation (i.e., migration to cloud), technology currency, AI applied to corporate systems like supply chain, and C\cybersecurity to name a few.

For companies to ensure responsible AI development and deployment, they must now consider their:

  • Energy demand and carbon emissions
  • E-waste/IT infrastructure end-of-life management and circularity
  • Water usage
  • Unfair/biased decision making and discriminatory outcomes (human impact)
  • Data security and customer privacy
  • Unethical business conduct
  • Equality’s Sustainability Principles for Enterprise Architecture will help EA organizations stand up new ways of assessing the viability of solutions against key sustainability principles while defining alignment to Governance, Environmental and Social industry standards.

Architects can leverage these principles in their Architecture Guidelines and Guardrails and apply them to business and IT initiatives. Reviewing solutions in terms of responsible AI and sustainable IT should be added as an EA capability similar to existing capabilities tied to tracking application rationalization, modernization to N-2, etc., as part of an Architecture Review Board. As part of their assessments, Architects have an opportunity to incorporate environmental, social and governance principles, standards and metrics that are being proposed as enterprise standards linked to federal and UN guidelines.

During an architecture review, Architects and engineers should consider the sustainable impacts (both positive and negative) that each option provides. Once a solution is agreed upon,

Architects can provide notice of decisions as to the degree of alignment to sustainable IT standards and track metrics that can be reported out to leadership.

AI intersects with sustainability/ESG in many ways. Successful organizations look at the intersection from a risks-and-opportunities perspective and tie their approach to responsible AI to their overarching sustainable business strategy. This will require organizational adjustments and alignments, as responsibilities for AI and sustainability span multiple functions that include IT and LOB personas. This means assessing sustainability as part of your Business Architecture artifacts.

As sustainability strategies become more holistic and ESG materiality more the driver for action, responsible AI and sustainability strategies need to account for the diversity in ESG issues that can be caused using AI, and practitioners must be able to pick the solutions that truly help address their organization’s most material issues.3

I found several research papers on Green EA or “Grean” as MDPI calls the practice.4 EA as a competency ensures three fundamental purposes: value creation, enterprise coherence, and strategic alignment. EA coordinates the roles, processes, information, applications, and technology necessary for an enterprise to fulfil its EA experience. As a practitioner for the last 17 years, I would break technology into integration, infrastructure and security and always start with the business and its needs as the front door before starting any work. EA’s role is to ensure enterprise coherence by integrating diverse resources into a unified, cooperative system. Strategic alignment, a pivotal aspect of EA, involves transforming strategic resource decisions into an organizational blueprint.

This blueprint channels resources towards value creation, in harmony with strategic goals and the overarching vision of the enterprise and alignment to CIO and CEO strategic drivers. For those working with or as an architect, this is not new news.

However, we do have an opportunity leveraging the need for digital transformation to incorporate ESG objectives by facilitating the strategic integration of sustainability aspects into the organization’s digital transformation as a way to push an organization’s ES agenda to the forefront.4 As many organizations are in the process of identification of where AI can help solve sustainability issues, Green EA will need to update its principles to include the encapsulation of responsible AI principles, patterns and practices within sustainability, technology and information principles, patterns and practices. AI and sustainable IT principles should be included in EA Architecture Review Board’s (ARB) notice of decisions and their corresponding metrics.

If we were to look at what is concerning top industry CIOs tied to sustainability, we can start to link architectural principles, standards and even “Green EA and Green coding efforts” to ensure that AI is being used to solve appropriate problems while ensuring the solution minimizes erosion of sustainability efforts.

The blueprint becomes a “transitional” current-to-future state North Star that must now take sustainable IT and AI principles, patterns and metrics of adherence into account. Organizations must embrace ESG for their long-term prosperity. But to take advantage of ESG opportunities and truly make a difference (not to mention complying with future reporting and regulation requirements)5 organizations must embed ESG goals into their business, data and technology.6 We can link responsible AI principles with sustainable IT principles and link both to existing EA principles and/or leverage them as new entities that must be adhered to while reviewing, selecting and deploying technology solutions.

If we now look at the sustainability model, we could take a template from IASA’s BTABoK ( and link ESG sustainable IT criteria to the highest level of the business model analysis framework including specific questions architecture must answer tied to the sustainable IT considerations in a PESTLE Canvas, Attachment 1. The PESTLE canvas provides and ecosystem view of the enterprise or an initiative. It is a comprehensive framework that examines six key areas: Political, Economic, Social, Technological, Legal, and Environmental. By conducting a PESTLE Analysis, businesses can better understand their current and potential markets, identify potential risks and opportunities, and make more informed decisions.

The Info-Tech Research group defined responsible AI guiding principles based on its goals, identified risks and other inputs. We can leverage this and include it within EA’s body of knowledge tied to principles and guidelines.

  1. Data Privacy: AI applications must respect user privacy. Data must not be used outside of agreed-upon terms and must be compliant with privacy norms and regulations.
  2. Accountability: We must have clear accountabilities (Incl: roles and responsibilities) assigned for all aspects of AI systems, including governance, incident response, and lifecycle management.
  3. Explainability & Transparency: AI applications will be transparent about how data is used and will provide users and key stakeholders insights into how outcomes are produced.
  4. Fairness & Bias Detection: AI applications must include checks and balances to ensure results are unbiased and there is fair and equitable representation across users.
  5. Security & Safety: AI applications must be resilient to attacks and other risks that could provide physical or digital detriment to individuals or groups.
  6. Validity & Reliability: AI applications must produce results that are accurate and consistent to mitigate AI risk and foster trust in the application.

There is an opportunity to create optimal AI solutions where architecture and engineering consider the linkage of AI principles and sustainable IT enterprise architecture principles together from an ESG perspective. EA should then update its broader principles and patterns where appropriate with specifics and metrics tied to sustainable IT and AI principles. This linkage would enable the business and IT to track their agreed-upon approach and solutions against key metrics. Each of the AI principles can be linked back to one or more EA principles. AI principles 1, 2, 3, 4 and 6 above all have ties to the EA principle of Ethical Foundation and its associated statement that sustainable architecture must be founded in ethics, i.e., data responsibility and privacy, digital equity and education, trust and security, explainability, transparency, inclusion, moral agency, value alignment, accountability, and hedging against technology misuse. AI Principle 4 aligns to the EA principle of Digital Inclusion and its corresponding statement that Information systems will be accessible to all, promoting digital inclusion. The principle of digital inclusion ensures that the benefits of technology are shared widely, supporting social sustainability and compliance with regulations such as the European Accessibility Act (EAA) or American WCAG. AI Principle 5 would need to be included in EA security and safety principles aligned with enterprise security capabilities, risk mitigation measures and security compliance.

The Architecture review would be like existing Architecture Review Boards where business and IT proposed AI solutions are reviewed leveraging recommended options with pros/cons. The architecture competency can capture specific metrics as well as alignment to architecture principals and patterns, noting notice of decisions. During selection as well as during delivery the architecture competency would capture specific metrics that would be reported back to IT and business leadership that are in line with business and strategic ESG initiatives. EA metrics are usually listed in four categories: (1) IT metrics, (2) customer metrics, (3) business/strategy metrics, and (4) compliance metrics. To ensure green EA, we would add a new fifth category with net new metrics tied to sustainable IT and responsible AI.

As an example under the category of emissions standards, EA would need to assess the viability of a solution against’s application portfolio emissions standards EF 213, EF 213-1 and EF 213-2. One method for expressing portfolio emissions is software carbon intensity (SCI). The factors used to determine SCI of a given application or suite of applications is E (the energy consumed by the software in kWh) X I (emissions associated with that amount of energy consumption) + M (carbon emissions of the hardware that the software runs on) per R (a functional unit such as end user or device). (Refer to Sustainable IT Standards Taxonomy V1.1 at

With the growing trend towards green EA and the rising value of AI, EA must adjust its scope, accountability and roles. LeanIX7 has produced a Business Capability Model for ESG that can be leveraged by EA and the business to understand areas of impact. See Attachment 2 below. As mentioned, architecture and business roles will need to transform as well.

Business Partners working with LOBs will need to work with Business or Business Domain Architects to ensure the opportunities and options being assessed consider ROI and alignment to achieve strategic and business objectives as well as ESG impacts that advance or erode a company’s ESG ambitions. Business Architectural Viewpoints will look for opportunities to take non- sustainable processes or application capabilities to green. If there is significant human capital to do processes that can be automated with AI, how do you ensure you are vetting the “right processes” to automate that maximize business value while improving efficiency metrics? Hence, option assessment should include alignment to responsible AI principles that are inclusive of AI automation impacts on sustainability.

Information Architecture and corresponding viewpoints must prioritize data security, access rights and usage. Consider optimizing data storage for energy efficiency and utilizing cloud technologies for resource efficiency, aligned with “Green data management” and standards. Consider “Green data management,” which aligns with AI sustainability (e.g., Standards Taxonomy V1.1 – Data Governance SIT G200. Implement efficient data management practices like deduplication and compression to reduce energy costs and adhere to legal requirements. Analyze data life cycles for improvements in data volume, lifespan and backup strategy. Assess the necessity of LLMs considering environmental impact.

Technology/Infrastructure Architecture and Engineering must weigh the environmental impact of AI technology against its societal and governance implications. The rapid expansion of AI leads to increased energy consumption and computational demands, necessitating significant infrastructure upgrades. This growth results in higher carbon emissions and the creation of physical assets with embedded carbon, requiring eventual decommissioning. Evaluating the adequacy of current technology infrastructure is crucial for supporting AI product generation. Mapping technology infrastructure helps identify ESG goals such as transitioning to low-carbon data centers and optimizing infrastructure usage to reduce emissions.6

The role of IT Vendor Management and Procurement is changing to accommodate new internal strategic directives that must be considered when selecting products and services. According to an IDC survey, one-third of IT buyer survey respondents said that they only work with or buy from IT vendors that meet certain environmental sustainability criteria.

Only 2% said that they do not make buying decisions that factor in environmental considerations.8 IT Vendor Management must be involved in Business Application and Technology Assessments, but their role is growing in scope as sustainability takes a front seat. Practitioners need to balance a technology’s environmental footprint against its potential social- and governance-related impact (“responsible AI”) when working with product and system integration vendors.

Implications to the Architecture Review Board

Modern day ARBs consist of architects representing different technology competencies, with representation from LOBs and IT Leadership as well as vendor management working in concert to rationalize priorities and alignment of viable options to deliver against corporate and business strategies. The modern-day ARB focused on providing IT Guidance and Guardrails now needs to include how to review business needs and technology options against the EA principles and directives we discussed in the previous section. This paper is specifically tied to AI but to meet and report out on sustainable IT measures the Green EA competency needs to become proficient in understating the SustainableIT standards taxonomy and how to effectively assess, research and vet technology/solution options for their ability to achieve positive business outcomes while ensuring they meet sustainability goals. Capturing the pros and cons of the recommended options and the decision records of what was voted on and agreed to by whom will be critical for traceability.

The ARB should identify critical stakeholders and define their roles as part of a standard RACI model tied to the decision-making processes and methods. ARBs will need to adjust their Decision Records: If we look at how Decision Records are documented and shared with IT and Business Executive Leadership, the architecture competency will need to update the traditional “reporting up” structure to include additional roles now being hired in many organizations. In the past, Decision Records went directly to the CIO, CDO and CISO as well as the Business Partner and executives owning the line of business. Now at least quarterly, if not monthly, updates would flow to the traditional IT leadership but include the Chief Compliance Officer, Chief Data Officer, Chief Risk Officer and potentially the C-level executive with ESG oversight.

While there are many opportunities yet to emerge, one thing is clear according to PWC: Enterprise Architecture Management is crucial for a sustainable IT strategy, where an ESG transformation in IT is a starting point. Enterprise Architecture Management guides IT landscape development and links technology with business activities. It not only helps clarify strategic issues but also provides full visibility down to a granularly level of IT artifacts and components.7

Holding the role of Chief Enterprise Architect for over 17 years, I feel this is a call to action for all EA competencies and VPs/Chief Enterprise Architects. We must ensure we are applying our Responsible AI and ESG EA principles and best practices to aid our business areas and executive leadership in determining sensible usage and deployment of AI. If not, our organizations will likely fail at achieving their ESG goals and exacerbate the carbon emissions problems we face today. AI technology can be a significant game changer in our society, but let’s ensure as EAs it’s used appropriately and not at the cost of a healthy climate!

1GreenEA”, Peter Klement, January 17, 2023. peter-klement/?trackingId=SaFvg0G1STKygW18BEUKZg%3D%3D

2 Why Green Coding is a Powerful Catalyst for Sustainability Initiatives, January 29,2023, IBM Cloud Education,

3 “AI Governance and Risk”, FY24, AI Governance: Info-Tech Research Group

4 Niels Vandeveene, Jonas Van Riel, Geert Poels, Green Enterprise Architecture (GREAN)- Leveraging EA for Digital Transformation, September 22,2023, 1050/15/19/14342#:~:text=The%20architecture%20layers%2C%20encompassing%20strategy,operation%2
0 blueprint%20of%20an%20enterprise.


6 Rim El Kadi, “An Enterprise Architecture Approach to ESG,” May10, 2022, ESG-PWC,

7 Sustainable Enterprise Architecture,”PWC, Viega, LeanIX, May 2023

8 Sustainable AI and AI for Sustainability, April 24, 2024- IDC ai-and-ai-for- sustainability/