Parallel Paths: ESG and Generative AI
The nature of modern business demands attention to both
It’s been a busy few weeks with lots of travel from DC to NYC to GreenBiz in Arizona. It feels good to be back home and writing again.
Two weeks ago, I was asked to join an in-person event and sit on a panel about AI and ESG. Of course, I am not an AI expert, but having a technology background, I know enough to be dangerous (that’s part of the ‘Benevolent Troublemaker’ in my LinkedIn headline).
As I sat and listened to the presentations and panels throughout the day, it amazed me how many similarities exist between ESG and Generative AI, not only at the surface level but down to the core and through to outcomes. In fact, by day’s end, I wasn’t sure which I had sat in a session for because you could easily swap out the topics with each other.
Admittedly, I see ESG in everything, but I don’t see Generative AI solving every problem. Yet, both are well poised as the risks and opportunities that surround business are changing. There are differences; one is a mindset/framework, and the other is a technology. But then again, read this statement:
With this in place, a company can drive more efficiency, be more effective in navigating interconnected challenges across datasets, and deliver long-term value for stakeholders.
Am I writing about ESG or Generative AI?
Let’s look at some of the parallels between the two and examine how they can solve some of a company’s most pressing issues together.
Confusion
There is a lot of confusion about what ESG and Generative AI are. Whenever someone talks or writes about it, it often requires an explanation. Since this is an ESG newsletter and I tend to stick with its original meaning while tackling the nuances, I assume we’re covered, but just in case:
ESG represents material risks and opportunities to the company that can affect long-term value.
Still, ESG is often open to interpretation because it can also intersect with ‘doing well by doing good,’ philanthropic endeavors, and saving the world and its people, but not always. As a result, its meaning has been twisted and warped to serve broad purposes, like aligning with stakeholder sentiment and values to capture a sale.
Generative AI has less nuance and the same issue with the perception of breadth.
Generative AI is an AI trained on mountains of unstructured data, including text, images, and videos, and can generate output from this information. It requires high-quality data, compute resources, and a model. While Generative AI handles unstructured and qualitative datasets very well to generate content, provide summarizations, create code, and more, Machine Learning is better suited for analyzing numbers, databases, and spreadsheets (so far).
A company can customize these models with its business or other data sets, which makes its commercial applications appear nearly endless. Still, these models aren’t a catch-all use case to solve every problem.
Like with ESG, Generative AI is sometimes seen as a solution in pursuit of a problem, resulting in a strategy-setting exercise that looks more like research.
Nevertheless, there are ways to solve a company’s most pressing issues with a thoughtful approach. And trust me, use cases are hanging out there waiting to be solved. We’ve been talking about Digital Transformation since around 2014. Still, a company shouldn’t just ‘add’ Generative AI into the mix and expect magic to happen, just like the value of ESG isn’t realized simply because you have a sustainability team.
ESG and Generative AI involve a mindset shift to understand that:
ESG involves pivoting your mindset around traditional financial metrics to interconnected ESG risks and opportunities in the context of your business.
Generative AI involves shifting your mindset around traditional applications and websites into natural language and the potential of data to address risks and create new opportunities in the context of your business.
Once the confusion is lifted, the execution looks consistent. Both are more powerful when used in concert with existing analogs. For example, ESG becomes actionable when combined with existing operational and financial data for analysis. Generative AI can be immensely powerful when supplemented with agents running quantitative analytics or augmented with automation, allowing people to chat with it inside a business application to complete a task.
Still, neither is a panacea, and indeed, the quality execution of either isn’t going to deliver alpha consistently and universally. I mean, what does? However, with quality Governance in play, both provide long-term value protection and creation when aligned with a business’s context.
As with many terms mired in confusion, there are those looking to capitalize on each. Be wary of anyone stating that there are no trade-offs, externalities, or issues. These initiatives take work. Without proper change management, the right skills matched with institutional knowledge, and leadership support, implementing these strategies could fail spectacularly.
Talent
While ESG has been around since 2004, it tipped over into mainstream attention around 2020. As you probably know, Generative AI is also a very recent development. With both, there are new requirements necessitating the need for training and talent development as not many specialists exist in the market. Indeed, even experienced industry experts will struggle in their verticals to find how each can integrate into the business. Microsoft (where I work) has free training for sustainability and AI because these skills intersect with technology and are relatively nascent. We need more skills in the market to progress.
While experts are needed, skills shouldn’t be held by a single group. Indeed, even those in non-core functions understand a business model ‘well enough.’ ESG and Generative AI have power when experts are embedded alongside core business groups. This union is one way to mainstream ESG throughout the company and find areas of differentiation, but this also applies to Generative AI. Pairing up technology experts with business expertise can spot the opportunities and yield more effective solutions...and possibly identify areas of technical debt that inhibit progress and opportunity.
Passionate people with core business knowledge are waiting for an opportunity to engage on these topics and grow their careers. Don’t miss the chance to capitalize on their passion by encouraging their skilling and growth.
While people are in the middle of ESG, inside the Social pillar, they are also at the center of any transformative initiative, including Generative AI.
Leadership attention
A common challenge may be senior leadership sponsorship and the external pressures leaders face. Across ESG and Generative AI, the management team hears things from the market, board, and investors. We’ve seen this with the rise of ESG on earnings calls and its recent subsequent decline. AI mentions are rising in earnings calls this year as the market seeks new places to uncover growth. A leadership team that doesn’t understand AI may do more harm than good.
Instead of saying the things and hoping an investor thinks you know what you’re doing, how about integrating these ideas into the business in a way that only you can? Unfortunately, this is easier said than done. These projects require funding and inertia is one of the most powerful forces in business. Capital allocation for ESG and Generative AI is a leap of faith because they both have unproven long-term value in mind.
The status quo just won’t cut it, though. Even if companies stand still, the world is changing around them, which is what ESG is all about.
Cutting through the market noise with talent that looks at these issues through a business context will help deliver focused business use cases and a story that will resonate with investors and stakeholders. To achieve this, leaders must be more aware of their business and trust their employees to understand these concepts and where they intersect with the company.
Today, business leaders are stuck thinking ESG is solely a compliance exercise, leaving risks unmanaged and opportunities on the table. Meanwhile, many are trying to capitalize on Generative AI, taking the leap on skunkworks projects instead of leading in small, provable use cases on their most material challenges.
As it turns out, the same tools that can help find ESG issues can find Generative AI use cases: a materiality matrix and stakeholder mapping. These tools can be the starting point to identify opportunities to solve what have long been unsolvable problems.
If leadership doesn’t foster innovation in these areas in the context of the business, we’ll see AI decline on earnings calls in about three years, as no one will have done anything. If you think ESG has gone away just because investors have stopped talking about it, I have news for you. ESG hasn’t gone away because it isn’t something the company generates; it comes from outside the business. Similarly, the competitive threat of Generative AI won’t go away either.
Overall, what leaders are missing is that they have long had the tools to value creation that could have set them up for ESG and Generative AI success. The few companies that have done this are already moving around it: Digital Transformation.
Data
As companies rush to disclose and report around ESG and companies rush to uncover their first killer Generative AI use case, there is one evident and foundational connection between the two. Data underpins both.
That doesn’t mean both are data, but one could make that argument. Many believe that ESG is data. For practitioners, ESG represents material risks and opportunities that can affect long-term value (in case you skipped the first section), as informed by data. As outcomes occur, the data changes and can reinform new outcomes. For analysts, ESG is data.
Is Generative AI data? Well, it generates content across new text, images, and more. Generative AI also does not exist without data—a lot of it. Generative AI almost does what ESG practitioners do: it uses the data to inform an outcome. The responses it generates likely are in service of a more immediate stakeholder need, unlike ESG. But like ESG, Generative AI can build long-term value.
With ESG on the decline in earnings calls, it may seem like leadership teams, at least in the US, have moved on to the next shiny thing. However, I have a secret for those reading, which I shared in a GreenBiz session last week.
Take heart if you are working in ESG or dealing with regulatory requirements! You have immense power to transform your company, but maybe not how you’ve been attempting it.
The same data exercises around ESG are required to stand up a quality Generative AI product, and this challenge isn’t going away.
ESG and Generative AI need governance practices and project work around data, such as:
Data inventory and identification
Data ingestion, cleaning, and transformation
Data science and analysis
Democratization of data across the workforce for more informed and timely decisions
As companies have organically fostered data and organizational siloes, they are in no position to deal with ESG or Generative AI projects, at least not at scale. If a company hasn’t done the table stakes to break down data siloes securely, lead with data governance, and democratize access to the data, they likely have more fundamental issues than they will admit. This, by the way, could be considered Digital Transformation. Welcome back to 2014!
It is at this point that ESG and Generative AI are the closest. The data and analysis through each provide an interconnected perspective to deal with interconnected risks and opportunities that continue to grow.
Again, nothing is a panacea, and humans must be involved here. Still, between an ESG mindset and new technology to deal with unstructured data, the company can build a much better defense to survive the short-term trends in pursuit of long-term resilience.
As this decade continues, data will remain a big theme, and Digital Transformation will progress, no matter what other corporate ideas emerge.
It’s time for action
One of the complaints I’ve heard from those working in ESG is that there is too much focus on the metrics and disclosures. Historically, I would have written that data brought together for reporting would be automated and analyzed by tools like Machine Learning and Generative AI, allowing sustainability teams to free up their time for action. This sounds like a fantastic proposition, but I’m not convinced companies are pursuing action.
While the data and Digital Transformation challenges are fundamental and will no doubt be critical to survive, until all of the pieces here are resolved, including understanding the value of ESG and Generative AI, those working in the space will continue to struggle to find sponsorship and budget to move.
On the heels of this challenge are others. Both ESG and Generative AI have backlashes for various reasons. While I didn’t cover this (I have backlash fatigue), aligning with the company’s purpose is the path forward.
There are also two conflicting perspectives on who can executive. First, there is a perception that only the world’s biggest companies can deal with ESG or Generative AI, but don’t believe it. Every company has an opportunity if they know where to look and is willing to invest. If you are in a big company, start with a small, focused project in one business unit before breaking down those data siloes. A win can carry your ideas far.
On the other hand, some believe only small companies are agile enough to progress. This one feels more likely to me, as startups have been a source of disruption for years now. If you are at a small company, try revisiting those use cases you couldn’t solve previously or consider how to augment what you're doing with ESG or Generative AI. From here, I’m sure you know how to innovate.
And so, there is hope to push through some of the inertia using the existing common opportunities. Who is courageous enough to try?