In most companies, ESG discussions typically focus on issues such as pollution control, biodiversity, health and safety, business ethics and boardroom diversity. Technology-related risks and opportunities do not receive adequate attention. But three areas where corporate managements need to start focusing on now are Green Software, AI bias and Trusted Data. In the future, they will have enormous implications in all the three – E, S and G – components of the organisation.
It has come to the forefront because of the pandemic-fuelled exponential increase in cloud adoption by global enterprises. Data centres already account for over 1% of the total global electricity consumption annually. That is expected to skyrocket and become 8% of total global electricity demand over the next 10 years.
Data centres not only consume a lot of electricity, they also require a lot of water to keep cool. The environmental footprint of data centres is becoming a big area of concern around the world. Optimising hardware and using solar or other renewable sources for electricity helps in reducing the carbon footprint to an extent. But an area that can also help immensely is green software – where the software’s algorithm ensures maximum energy efficiency. This is critical because the electricity consumed in data centres is directly dependant on how efficiently software applications handle hardware resources.
In simulations conducted at the University of Washington, green software development techniques reduced energy consumption by up to 50%. Earlier this year, the Green Software Foundation – founded by corporates and non-profits including as Microsoft and Linux Foundation – took on the task of mainstreaming the sustainable coding movement. It is currently in the process of establishing green software standards and practices across various computing disciplines and technology domains. Looking ahead, sustainability officers would want to ensure that the software developed by their employees and vendors includes green practices which are subject to energy monitoring, peer benchmarking and performance reviews.
As corporations increasingly harness the power of artificial intelligence in everything from recruitment decisions to customer care, concerns related to AI-bias are also being flagged. Algorithmic or AI bias can have profound implications in almost any area of deployment. For instance, this bias could lead to discrimination against minorities and women, and raise questions over privacy, especially around how much data is necessary collected to make decisions. If AI is used to make decisions about people that might cause undesirable impact, how are enterprises governing that? How much information about people is it appropriate to capture? What decisions are we going to let a machine make? All this could end up in a bigger social governance issue. For example, a large conglomerate recently apologized over an “unacceptable error” in which its AI-driven algorithms categorized a video about members of a minority community as being about primates. Companies need a plan for mitigating such risks. In order to ensure social equity, it is critical to have strong governance controls for developing and deploying AI solutions.
Currently, investors rely on two primary information sources for making funding decisions. The first is a company’s self-reported quantitative and qualitative data around ESG impact. The second is peer benchmarking of a company’s ESG performance, for which third-party ESG ratings are leveraged. Unfortunately, the plethora of rating methodologies often impedes objective decision-making. This issue can be addressed by triangulating the above data points by leveraging Natural Language Processing (NLP) techniques which can help conduct sentiment analyses on stakeholder perceptions regarding a company’s ESG policies and practices. This involves scanning online news and social media posts for positive attributes, as well as controversies, complaints, and potential legal actions. NLP enables the real-time conversion of millions of such structured and unstructured pieces of information – including text, images, and videos – into a structured and intelligent dashboard that can help “unify” disparate metrics. This can be aligned to various ESG frameworks and performance standards, and ultimately, used by investors to make more informed decisions.
Investors can also benefit from technologies such as blockchain which enable trusted and standardized ESG data collection and reporting. For example, a global phone manufacturer is using this technology to trace the origin of raw materials and work-in-process inventory across its global supply chain. It has a real-time view of compliance certifications related to quality, along with labour and environmental clearances. This facilitates transparency and auditability across the ecosystem of suppliers, subcontractors, distributors, and service vendors – which is what investors are looking for when they evaluate companies’ ESG practices and impact.
Organisations operate within a complex ecological system and ESG parameters serve as a proxy for the quality and impact of their interactions with diverse stakeholders. In this regard, leveraging emerging technologies can help companies create transformational solutions for addressing the ESG challenges we face today.
The writer is Technology Sector Leader and Chaitanya Kalia is the Climate Change and Sustainability Services Leader at EY-India.