Digital ecology and green artificial intelligence: concept, assessment methodology, and prospects for sustainable development

The purpose of the study is to provide a scientific substantiation and systematization of the concept of digital ecology and green artificial intelligence as a direction of sustainable development of the digital economy, as well as to develop an approach to assessing the impact of environmental AI indicators on the market value of equity of high-technology companies.

Methods

The methodological basis of the study consists of systemic, comparative-analytical, institutional and value-based approaches. The research uses methods of analysis and synthesis of scientific literature on Green AI, comparative analysis of corporate environmental reporting of technology companies, classification of digital ecology indicators, expert-analytical interpretation of ESG and climate factors, as well as financial modeling methods. To assess the impact of green AI on company value, the discounted cash flow method is applied, supplemented by an analysis of factors affecting revenue, operating expenses, capital expenditures and the discount rate.

Results

Based on the study, it is shown that green artificial intelligence emerged as a response to the growth of computational costs in machine learning and to the need to move from evaluating models solely by accuracy to a comprehensive assessment of their resource efficiency. The key groups of digital ecology indicators are identified: operational metrics, including energy consumption, PUE, utilization of computing capacity and energy per request; environmental metrics, including CO₂ equivalent, water footprint, resource depletion and the share of low-carbon electricity; and useful-effect metrics that characterize the contribution of AI to reducing emissions and improving the efficiency of energy, transport, industry and environmental monitoring.

Global green AI practices are summarized, including energy-efficient models, data center optimization, disclosure of environmental indicators and the use of artificial intelligence for climate-related tasks. Russian cases related to tools for assessing the carbon footprint of machine learning and the use of neural networks in environmental monitoring are considered.

Mechanisms for assessing the impact of green AI on the value of a high-technology company are developed through reducing the cost of inference, lowering ESG and regulatory risks, improving technological efficiency and increasing the trust of corporate customers. Using Mistral AI as an example, it is shown that taking environmental factors into account in the discounted cash flow method can change the result of equity valuation by adjusting cash flows, capital expenditures and the cost of equity.

Main conclusions

Digital ecology and green artificial intelligence are moving from the sphere of a declarative ESG agenda into the area of measurable engineering and financial-economic solutions. The environmental footprint of AI is becoming an independent object of management. The practical significance of the results lies in the possibility of using the proposed approach to analyze corporate reporting, assess the effectiveness of measures aimed at reducing the resource intensity of AI and substantiate managerial decisions in the field of sustainable digital development.