Back to podcast

Pallavi: Welcome to a new episode of EY India Insights podcast in ‘Generative AI’ podcast series. I am Pallavi, your host, and today we have an exciting topic to explore – How is Generative AI transforming ESG strategies of companies?

With innovative tools like ChatGPT and Dall- E, Generative AI has captured global attention. As it revolutionizes various sectors, it is now making its mark in the realm of ESG. To shed light on the immense potential of Generative AI in driving sustainability initiatives, we are joined by a distinguished guest, Alexy Thomas, Technology Consulting Partner at EY India. With an impressive track record spanning over two decades,  Alexy leads the data and analytics practice, spearheading the transformative efforts in global financial services sector. His expertise lies in data and analytics in ESG management, technology strategy, and program delivery, which enables organizations to embrace agile engineering for exceptional customer value. Welcome to our podcast, Alexy.

Alexy: Thanks, Pallavi. The potential of Generative AI is immense, and I am really happy to be here to discuss its relevance to ESG strategies.

Pallavi: While there is a lot of buzz about Generative AI (Gen AI) today, how will it help in ESG strategies more than the other AI versions, which too are based on machine learning or deep learning?

Alexy: Enterprises need very granular and accurate data and insights from a wide variety of sources to effectively manage their ESG strategy. Data required is often unstructured, like text documents, audio, and video files. Gen AI, powered by large language models (LLMs), can excel compared to traditional AI applications in tasks such as recognizing images, processing text, processing audio and video content. 

Consider a company that is looking to build a new manufacturing unit and needs insights about locations and the potential ecological impact of building in that location. It is a prime candidate for Gen AI capabilities that can be utilized along with other AI models to collect relevant data, including aerial footage. The data then needs to be analyzed with geospatial and open data to extract insights and to understand an impact on biodiversity, ecosystem sensitivity, and air and water quality implications. This will enable the company to make better, more informed decisions. It is important to recognize that it is not one technique or one model, but a collection of models and techniques that work in ensemble, along with other advanced analytical techniques and     Gen AI in a complementary way to implement these solutions. 

Gen AI, with its ability to analyze and interpret data, apply logical rules and constraints, and infer new knowledge from existing knowledge, is a key component to these emerging ESG solutions.

Pallavi: The different AI programs and applications from ChatGPT and DALL-E are these specialized and focused solutions?

Alexy: The last few months have been really exciting, and a number of Generative AI models and systems have emerged like GPT-3, GPT-4, LLaMA, BERT and DALL-E. All these are foundational models trained on large amounts of data and have billions of parameters. By building on top of these foundational systems, we can create more specialized and sophisticated models tailored to specific use cases and domains. Being pre-trained on massive amounts of data, these foundational models deliver huge acceleration in the whole AI development lifecycle, allowing businesses to focus on fine tuning to their specific use cases. Choosing the correct LLM model to use for a specific job requires expertise, and enterprises need to deploy specific skills to fine tune these models for ESG and for their specific context and sector to ensure these models provide meaningful insights. Training these tools on inaccurate or skewed data will lead to generating false or biased insights. Therefore, it is necessary to build in the right controls and governance. 

To summarize, foundational models are great starting points but need to be fine-tuned for ESG to the specific context of the enterprise to get the best results.

Pallavi: How does Generative AI help in managing emissions, net zero ambitions and biodiversity, and any other such issues? 

Alexy: The best way to answer that question is by sharing a couple of examples on how enterprises can use Gen AI techniques to get better insights and to manage emissions.

Let us say a large multinational retailer wants to streamline its processes for collecting greenhouse gases (GHG) Scope 3 emission data. Gen AI tools can help to analyze the data and derive insights to improve supply selection and ratings. It can provide automated personalized guidance to supply chain partners, thereby helping improve their carbon footprint. 

Another example is from a regulatory and disclosure standpoint. Globally, ESG policies specific to countries and sectors are still evolving. Reporting rules differ widely, which can be a huge challenge for enterprises. LLMs can provide insights and guidance on ESG regulations specific to the regions and countries that these enterprises operate in, which can help companies understand these different policies and manage them accordingly. I see these Gen AI-powered co-pilots increasingly being used as significant enablers in managing and enabling ESG transformations and the transition to net zero.

Pallavi: This has been a very interesting discussion, and I have learned a lot from it. Thank you for sparing your valuable time and joining us for this podcast.

Alexy: It was a pleasure and thanks for having me. 

Pallavi: On that note, we come to end of this episode. If you would like us to explore other such topics on Generative AI, please do leave us some suggestions that you would like us to deep dive into. Thanks for listening in, and goodbye for now.

Back to podcast