At Rise of AI 2023 in Berlin, Dilek Sezgun, AI Techsales Leader at IBM, argues that AI is essential for companies to handle ESG (environmental, social, governance) work, which she frames as covering far more than emissions, including employee expectations, company structure and supply chains. Citing IBM Institute for Business Value research, she says 60 percent of CEOs already track their sustainability transparency, 64 percent know their sustainability goals, and 80 percent are investing in the area, while a 2021 study found 42 percent of CEOs recognize digital technology as important for sustainability. She presents IBM's model for turning sustainability ambitions into action across strategy, ESG reporting, infrastructure and supply chain, built on a data foundation she calls Cloud Pak for Data, AI models including Watson Discovery for extracting ESG data from unstructured text, and reporting and visualization tools including Cognos, Envizi and OpenPages for compliance. She identifies siloed, inconsistent and untrustworthy data as the biggest recurring obstacle, summarized as garbage in, garbage out, and closes by pointing to a six-week MVP approach IBM uses with clients, including a European bank and the energy company E.ON, and a workshop deep dive later in the program.
Key takeaways
- Sezgun cites IBM research finding 80 percent of CEOs are already investing in sustainability, though only 42 percent (2021 study) linked it to digital technology.
- She frames ESG as covering employee expectations and supply chain, not just emissions, with companies judged on all three dimensions.
- She names siloed, inconsistent and untrustworthy data, summarized as garbage in, garbage out, as the biggest recurring obstacle to ESG reporting.
- IBM's Watson Discovery is used to extract ESG data from unstructured data sources for reporting.
- IBM builds sustainability proofs of concept with clients in six-week 'client engineering' sprints, citing a European bank and the energy company E.ON as examples.
Detailed notes
Why Sustainability Is Now an AI Problem
Sezgun opens by framing AI and sustainability as two forces fundamentally changing industry and the world, arguing ESG, environmental, social and governance, goes far beyond emissions to cover which employees want to work for a company, its internal structure and its supply chain, and that every company will be judged on all three dimensions. Citing IBM Institute for Business Value studies, she says 60 percent of CEOs already know their sustainability transparency level, 64 percent know their sustainability goals, 80 percent are already investing in sustainability, and, from a 2021 study, 42 percent of CEOs recognize digital technology as important for achieving it. For 2024 she identifies three priorities: ESG reporting, including data, integration, governance and benchmarking; the operational side, meaning skills, metrics and supply chain; and green IT, meaning the infrastructure sustainability runs on.
IBM's Sustainability Model and the Data Challenge
She presents IBM's approach as turning sustainability ambitions into action across strategy, organization, ESG reporting, facilities, infrastructure and supply chain, built on a data and AI foundation she calls Cloud Pak for Data, which lets companies share data assets under regulatory compliance. She illustrates this with an MVP built in a six-week 'client engineering' effort for a European bank, using EU taxonomy data. She names the biggest recurring challenge as data itself: siloed data sets with no connections between them, pressure from CEOs and CIOs for quick sustainability answers, lack of standardization, and lacking accuracy, which she summarizes as garbage in, garbage out, meaning trustworthiness of input data determines the quality of any sustainability reporting.
AI Models and Reporting Tools
For turning unstructured data into usable ESG information, she highlights natural language processing, specifically Watson Discovery, which extracts ESG data from various data formats and interfaces with other software. For reporting, she describes a tool, to be demonstrated later in a hands-on workshop, that needs only a small number of data sets to analyze supplier CO2 footprints, run predictions, and let a company balance CO2 reduction goals against cost. For visualization and compliance she names Cognos for dashboards and sustainability reports, Envizi for reporting with regulatory requirements built in, and OpenPages for capturing ESG priorities, strategic objectives, process risks, controls and indicators to keep reporting compliant.
Q&A Highlights
Asked where clients tend to get stuck first, Sezgun says most CEOs already know they need to act on sustainability; the real obstacle is operational, not knowing which data to use or which reporting approach to take, which is where IBM helps through partners, consulting or building an MVP or proof of concept. Asked whether particular sectors are hit hardest, she says every sector faces its own version of the challenge, citing examples spanning the automotive industry to energy, including a reference case with the energy company E.ON built on IBM's platform.
Quotes
It's garbage in, garbage out.
Dilek Sezgun
Each company will be judged on that, how they are focusing on these three subjects.
Dilek Sezgun
It's not only about compliance, it's also the possibility for each company to grow, to innovate and to be better than others.
Dilek Sezgun
Questions answered in this talk
How many CEOs are investing in sustainability according to IBM research?
According to Dilek Sezgun at Rise of AI 2023, IBM Institute for Business Value studies found 80 percent of CEOs are already investing in sustainability, 64 percent know their sustainability goals, and 60 percent track their sustainability transparency.
What is the biggest challenge companies face with ESG reporting?
According to Dilek Sezgun at Rise of AI 2023, the biggest challenge is data: siloed data sets with no connections between them, lack of standardization, and lacking accuracy, summarized as garbage in, garbage out.
What IBM tools support ESG reporting and compliance?
According to Dilek Sezgun at Rise of AI 2023, IBM uses Cloud Pak for Data as a data foundation, Watson Discovery to extract ESG data from unstructured sources, and Cognos, Envizi and OpenPages for visualization, reporting and compliance.