Skip to content
Environmental

ESG Case study: How EnerSys uses GenAI to drive efficiency, ensure accuracy, and safeguard sustainability & ESG data

Natalie Runyon  Director / ESG content / Thomson Reuters Institute

· 6 minute read

Natalie Runyon  Director / ESG content / Thomson Reuters Institute

· 6 minute read

AI is revolutionizing sustainability practices, with EnerSys showcasing how AI tools can significantly enhance efficiency and accuracy in data collection, reporting, and analysis for ESG efforts

In the recently published Thomson Reuters Future of Professionals 2024 report, more than three-quarters (77%) of professional services respondents said they believe artificial intelligence (AI) will have a high or transformational impact on their work over the next five years. This was 10 percentage points higher than in the 2023 report; and moreover, a resounding 78% of professionals said they believe AI is a force for good.

It appears that this may be the case for sustainability practitioners as they face intense workloads from the explosion of upcoming environmental, social & governance (ESG) regulatory requirements. Christina Sivulka, the global sustainability manager at EnerSys — a leading industrial battery manufacturing and energy storage company — is one of these leaders at the forefront of leveraging AI. In fact, she and her colleagues have been using AI to enhance their company’s sustainability data collection and reporting processes for the last 18 months.

Their innovative approach first started with collecting Scope 1 and Scope 2 emissions and resource consumption data. EnerSys uses a platform called ESG Flo, which employs heat map-based machine learning to extract key information from utility bills at their 180 sites worldwide. This AI-powered system has significantly improved data accuracy, auditability, and efficiency in collecting Scope 1 and 2 emissions data; and, according to Sivulka, contacts at all 180 sites “upload PDFs of their utility bills, which the AI then processes to extract data such as date range, usage amount, cost, and units of measurement. The AI also flags anomalies and variabilities, which has been instrumental in helping us collect our data in a traceable and auditable way.”

The sustainability team at EnerSys is also piloting ESG Flo’s compliance platform, which contains project management tools supplemented by AI to help companies meet upcoming ESG regulations and disclosure requirements. The platform uses AI to populate answers to similar disclosure questions across different ESG frameworks, saving time and effort.


ESG Case Study
Christina Sivulka

[Once contracts are uploaded] the AI then extracts data such as date range, usage amount, cost, and units of measurement. The AI also flags anomalies and variabilities, which has been instrumental in helping us collect our data in a traceable and auditable way.”


In addition, EnerSys is using ChatGPT Enterprise to analyze large datasets related to sustainability metrics, including Scope 1 and 2 emissions, travel data, and waste data. This generative AI (GenAI) tool is helping the team members uncover insights more quickly than manual analysis. They are also using it to assist in answering customer questionnaires and surveys about their sustainability practices. By uploading EnerSys sustainability reports and internal policies to the ChatGPT Enterprise platform, the team can generate responses to customer inquiries more efficiently, although human review is still necessary to ensure accuracy. Sivulka estimates that the tool has cut the time spent on questionnaires by roughly 50%.

Looking ahead, EnerSys plans to explore using ChatGPT Enterprise to help write portions of the company’s next sustainability report and customize storytelling for various types of stakeholders. The team also is considering using AI to review its Climate Disclosure Project (CDP) questionnaire responses for potential improvements and gap identification.

Actions to overcome issues of trust and accuracy

With trust and accurate data continuing to be key concerns for using GenAI, Sivulka says that she and her colleagues are proactively addressed these issues through a couple key steps:

Prioritize internal collaboration from the beginning — EnerSys took a collaborative, cross-functional approach to address concerns and implement AI tools for sustainability purposes. When exploring the use of AI, they partnered with key stakeholders including the company’s IT, legal, internal audit, and compliance teams to thoroughly evaluate risks and establish proper controls. This allowed the company to customize cybersecurity and data privacy safeguards specific to their use cases. For example, EnerSys coded its ChatGPT Enterprise implementation to flag and reject requests involving proprietary or material information.

Provide training to employees — Employees utilizing AI at EnerSys received comprehensive training on the responsible use of GenAI tools like ChatGPT Enterprise. The training focuses on several key areas, such as how to effectively engineer prompts, learning cybersecurity best practices and data privacy protocols, and how to identify potential biases or inaccuracies in AI-generated content.

Further, employees were taught to always have a human review any AI outputs before using the data externally; and they were instructed on what types of information should not be entered into the system in order to protect proprietary data. Employees had to sign documents acknowledging their understanding of proper AI usage guidelines, while ongoing collaboration with IT, legal, and compliance teams helps the organization utilize AI safely and ethically.

Guidance to get started

Overall, Sivulka says that her advice to her peers at other companies is to leverage AI cautiously but proactively to stay ahead of the curve, as sustainability teams will likely need to adopt these tools to keep up with increasing reporting and data collection demands. Specifically, sustainability team leaders would be wise to:

Partner across the enterprise — Sivulka emphasizes the importance of cross-functional collaboration when exploring AI for sustainability purposes. She advises against operating in silos and recommends partnering closely with risk functions to scope out feasibility, mitigate risks, and set up AI tools safely. This collaboration was critical for her team in implementing AI solutions like ESG Flo and ChatGPT Enterprise among her team.

Engage actively with software vendors — Sivulka encourages sustainability professionals not to be afraid to talk with AI software providers, even if the professionals don’t have a strong technical background in AI or coding. Many tools are designed to be user-friendly for non-technical users, and AI software providers are often eager to discuss new ways AI can help individual businesses and address specialized needs.

Make human review mandatory — Sivulka stresses the importance of treating AI tools like ChatGPT as you would a human employee, particularly when it comes to reviewing work. She notes that humans are not always accurate and can have biases, so AI outputs require similar scrutiny. “I think you just have to be aware of it,” Sivulka says. “We have been trained to know how to flag or how to see any sort of bias or lack of accuracy.” She advises having humans review all AI-generated content before using it, in the same way as managers review their employees’ work.

As AI continues to revolutionize sustainability practices, companies like EnerSys are at the forefront of this technological integration. By leveraging AI tools for data collection, reporting, and analysis, EnerSys demonstrates how these technologies can significantly enhance efficiency and accuracy in an organization’s sustainability efforts.


You can read more ESG Case Studies here.