By Srinath Sridhar, Principal Consultant, Sydney
For most organisations today, as the business evolves, there is an increasing focus on their environmental, social and governance (ESG) pillars. However, for the most part, ESG in organisations today is limited to fulfilling their reporting obligations to the regulators. This is not sufficient to make a difference. Organisations need to move from passively reporting after the fact to actively moving the ESG dial by influencing the sustainability future using AI and ML.
With the growth of AI in general and Gen AI in particular, the number of use cases for ESG are going to skyrocket. We are at a pivotal moment where the rapid rise of AI and the urgency of ESG initiatives are going to combine to generate rapid progress for AI in ESG
Definition and Use cases of AI for ESG:
There are several intriguing use cases that come to mind for each of the three pillars of ESG (Environmental, Social and Governance). Let us evaluate key industries to understand key use cases to use AI on ESG.
A. Environmental: McKinsey defines the E in ESG, environmental criteria, includes the energy your company takes in and the waste it discharges, the resources it needs, and the consequences for living beings as a result. Not least, E encompasses carbon emissions and climate change. Every company uses energy and resources; every company affects, and is affected by, the environment.
- Retail and Consumer goods: For retail and consumer goods organisations, there is a steady increase of manufacturing and stocking of ‘green’ products. Green products are stock keeping units (SKU’s) that have a lower environmental footprint. Certain customer segments are willing to pay premiums for green products. A retailer can therefore use targeted marketing techniques to push these green products to such customers. In store planogramming, retailers can use AI techniques to exhibit the green products at key spots in the store. For online stores, retailers can feature green products prominently in the recommended products section. Consumer goods companies need AI for understanding pricing strategies for these products.
- Utilities: With the advent of IoT devices, utilities can understand and control wastage of precious natural commodities. Many water utilities have set up IoT devices in the water network and use AI to identify and curtail water leakage. Electricity utilities are increasingly offering tips for optimal usage of energy using smart grids. These AI use cases are leading to better conservation of critical resources.
Utility organisations are also extremely asset heavy. These organisations are using AI for predictive and preventative maintenance. This is increasing the life span of their existing assets and thus reducing the amount of assets sent to landfills.
- Energy: Energy organisations that are in a moment of great flux due to the increasing focus on renewable energies. The worldwide consumption of energy is unabated due to increasing urbanisation and world events. This is leading to extremely high carbon emission levels. AI can help energy companies with better and cleaner production processes. With the increasing migration to renewables, AI can also help with the balance of traditional and renewable sources of energy by understanding data from smart grids.
- Transportation and Logistics: It is a no brainer for transportation and logistics companies to move their existing fleet of vehicles to environmentally friendly options. However, these organisations are using sophisticated AI based scheduling and optimisation algorithms to plan their routing of vehicles. Total energy consumption (And cost) is therefore optimised leading to an environmentally conscious company.
B. Social: According to McKinsey S, social criteria, addresses the relationships your company has and the reputation it fosters with people and institutions in the communities where you do business. This includes labor relations and diversity and inclusion. Every company operates within a broader, diverse society.
- Universities: Post covid, it has been a very different way of working for universities. Students have moved to a hybrid model of education. Therefore, there needs to be much greater focus from universities on understanding of student wellbeing. There are different indicators of well being including academic results, engagement with the university, using of infrastructure like libraries and other aspects. AI and ML can play a predominant role in improving the social wellbeing of the university students by understanding these indicators and creating custom plans for student development.
- Financial Services: As the cost of living and inflation increases around the globe, this has led to a lot of repercussions for the financial services industry. Banks and Insurance companies are seeing a higher portion of their customers defaulting on their mortgage or other payments. Financial institutions need to engage on a humane basis with these customers. Instead of fobbing off a missed payment to a collection agency to deal with delinquencies, they need to take a nuanced approach that offers customers alternative options including payment plans. AI plays a significant part in this as it helps the financial institutions categorise customers into different segments based on their ability to continue their payments. Identifying potential defaults before they occur and having an alternate strategy for customers whose financial pain is temporary will ensure that the bank can balance its financial goals with social goals.
- Not for Profit sector: The not-for-profit sector is engaged on several initiatives for social good. There are organisations helping with the homeless, with children and women welfare, with caring for abused people, dealing with addiction etc. There are several opportunities to use AI across this sector to help with use cases including classification and segmentation for interventions, evaluation of charities and scorecards, using natural language algorithms for early detection of abuse worldwide etc.
- Labour intensive industries: For organisations that employ a workforce, there is a lot of focus on a happy and well rested workforce. HR Analytics provides an understanding of the end-to-end employee experience from hiring to resignation. It also can predict which employees are most likely to quit. There is also the workforce optimisation use case for industries that have a lot of contractor employees. Workforce optimisation can create an optimum working schedules based on resources available and priority of actions.
C. Governance: According to McKinsey G, governance, is the internal system of practices, controls, and procedures your company adopts to govern itself, make effective decisions, comply with the law, and meet the needs of external stakeholders. Every company, which is itself a legal creation, requires governance.
- Financial Services: As seen by the Royal commission in Australia, financial services industry must deal with, interpret, and implement a wide variety of governance and regulations. The Council of Financial Regulators (CFR) is the coordinating body for Australia’s main financial regulatory agencies. It includes the Reserve Bank of Australia (RBA), the Australian Prudential Regulation Authority (APRA), the Australian Securities and Investments Commission (ASIC) and the Australian Treasury. Most of the problems are analysed after the regulator has identified the issue. AI is required to predict where the governance lapse will happen, rather than post facto after the regulator has pulled up the organisation.
- Knowledge based industries: In knowledge-based industries such as law, regulatory and others, extensive domain knowledge is required to be fully abreast of the vast information available. With the fast-moving nature of the industry, a new category of consulting organisations has sprung up to assist in the collation and dissemination of this knowledge. AI plays a major role for these organisations to scrape data from various and websites as well as being equipped with an NLP based search and retrieval mechanism for knowledge retrieval.
Data management and Governance to enable AI for ESG:
It goes without saying the sound data management and governance is required to enable ESG AI use cases. However, there are a few specific aspects of how ESG feeds into the data to insights lifecycle.
- Data engineering: Let’s take the example of the Electrical utilities sector. In the power utilities world, 5-minute settlement is an example of a regulatory change that has been adopted industry wide. 5-minute settlement is to bring into line the generators and the settlement of electricity prices. This leads to higher volume of data as well as higher data frequency.
- Data governance and management: For all use cases of ESG, the governance, lineage and easy cataloguing of the data remains central to the use case. Given the implications of reporting incorrect data to the regulatory bodies, this is an area that needs strong oversight.
- Data privacy and security: The last few years in Australia has seen a plethora of examples of data security issues- with private data being compromised and leaked to the public domain. Sound data security is required for both data at rest and data in motion to ensure that organisations can safeguard critical customer data, increasing the overall governance of the organisation.
- Data modelling: ESG is very domain centric in nature. Within every industry, there is a lot of reusability of how ESG AI use cases are built and consumed. Because of this, there is a lot of merit in creating industry specific data models for all ESG consumption use cases.
ESG and AI are together whipping up the perfect storm. We expect to see rapid movement in this space, so continue tuning in.
For more information on how Altis Consulting has helped organisations implement ESG AI use cases please connect with us here.