As regulation increases, companies are looking to generative AI to simplify ESG and GRC reporting

The raft of new global legislation for financial, GRC and ESG reporting is putting new pressure on companies to consolidate and clean up their data and processes. Could generative AI alleviate some of this pressure? Companies aren’t sure how best to make use of it, and the regulations themselves are still evolving.

This was on the minds of delegates at last week’s Amplify EMEA conference in Amsterdam, organized by Workiva, which provides a cloud-based platform for integrated reporting across ESG, financial reporting and GRC (Governance, Risk and Compliance). The agenda highlighted trends for a rethink of data, reporting and governance. In this complicated market, where a mix of consultants, auditors and suppliers all work closely with their clients, AI is starting to make an impact – especially in the ESG space, where data sources and processes are not as predictable as for financial reporting.

Speaking on a panel entitled ‘Can you anticipate regulatory changes?’ Andie Wood, VP Regulatory Strategy at Workiva, told delegates that even regulators are now trying to streamline the increasingly complex demands on organisations. She says:

Many new regulations have been introduced over the past five years, and regulators are hearing from companies that they feel under enormous pressure as a result. I’m not just talking about one big reporting regulation, but about a whole set of rules for the Green Deal and for other changes in the environment.

We hear them talk about focusing on implementation. What are we going to do next? In fact, we saw that the commissioners were all instructed to specifically look to reduce the reporting burden. We are therefore very curious about what they actually think about a reduction in the reporting burden. Everyone gathers themselves. They decide what their new policy will be.

Undoubtedly, robust data and processes will play a major role in the new regulatory frameworks, said panel moderator Paul Dickinson, founder of CDP. He notes:

We’ve talked before about how if you really want to take full advantage of the almost limitless, almost unimaginable possibilities of machine intelligence, artificial intelligence, you need incredibly good data benchmarks now in order to benefit from that in the future.

The growing role of generative AI

Especially in the ESG field, having robust data and processes is an important step in allowing generative AI to play an important role. Naval Khana, director at PwC UK, says the technology can help companies meet their obligations under updated regulations such as CSDR and ISSB.

He presented figures at the conference from this year’s annual PwC CEO report, which showed that 68% of CEOs think generative AI will increase employee efficiency, and 44% expect it to increase profits. In the report, PwC says it is currently actively engaged with 950 of its top 1,000 US consulting clients, and is now discussing its use and implementation with its audit clients. Khana says:

Everyone has access to generative AI, and this has led to an unusual situation, a democratization of a new technology, where personal use of gene AI, such as ChatGPT, has come before companies were aware. It is unique and has a huge impact, which companies are watching with a mixture of excitement and fear.

According to Khana, generative AI is already generating content and analyzing and interpreting it. He says trends such as interpreting data from multiple media inputs using AI agents that can handle text, video and audio inputs will have a big impact on efficiency. Domain specific models such as Harvey in the legal arenaand AI systems that use ‘Human in the Loop’ models are also becoming increasingly effective.

However, Jeannette Gorzala, VP European AI Forum, told delegates that they should guide the technology until they are confident they can use the models and reduce risk anxiety. She says:

Two-thirds of employees now use ChatGPT secretly at work, but only a third of companies offer their employees advice on how to best use gen AI. So there is a gap, a kind of limbo, and if we want the companies to be able to overcome the pain points and innovate using gen AI, we need to have a top-down approach, in addition to the current bottom-up approach.

Gorzala, who represented more than 2,000 European AI companies during the EU AI Act legislative process, says the approach the law takes will help organizations scope their systems based on risk. She notes:

Gen AI will be disruptive and there will be pain points in reporting, but data literacy will gradually develop and the use of gen AI tools in specific areas will increase, without adding any risk.

Different paths to adoption

But while these companies may need the efficiencies that generative AI-based platforms provide, there’s no point in building their own generative AI systems from scratch, according to Simon Atherton, director of CFO Advisory Practice at Deloitte UK. He believes it makes more sense for organizations to leverage software companies like Workiva that integrate generative AI into their platforms. He says:

In certain areas, such as the front office, we are all working with organizations that are already using generational AI. But from a financial and GRC perspective, I think people are just trying to understand what the use cases are. My view is that most will never actually buy and build their own generation of AI applications, but will use them when companies like Workiva integrate them onto their platforms.

He draws a parallel with the rise of Robotic Process Automation (RPA):

About 5 years ago there was a huge trend for RPA software. Everyone started using RPA. I actually don’t think many people in the financial world use RPA because how you create reports, how you prepare invoices, how you pay invoices, how you maintain journals all need the same organization.

So I think vendors will build gen AI into their platforms. That’s how finance works. When Workiva builds something that uses generational AI, it’s all about risk. Workiva will say, “We tested this. We have these protocols. This is it, otherwise you have to build it yourself.’ There’s no point in building it yourself.

Workiva’s Wood says organizations are currently unsure about incorporating generative AI into their reporting processes, but it has become a learning process. She notes:

There is uncertainty. To what extent can AI really draw? Does it provide high-quality disclosure? Does this lead to a common disclosure? But we definitely see it being used to outline the initial outline and then outline the circumstances. I think it’s certainly helping already, but we’re still learning.

Gorzala of the European AI Forum believes that it will be crucial for companies to take their employees on that learning journey if they want to use generative AI effectively and safely to meet their regulatory obligations. She says:

The disruption that AI brings will lead to innovation, offer great potential to achieve more and bring benefits to organizations. Including employees on the learning journey is an important first step. There will be huge real-time savings in automating the reporting of real-time information sources and using next generation AI tools to ensure compliance.

My opinion

Delegates at the Workiva Amplify EMEA conference are fed up with the ever-changing global regulations affecting them. And while financial reporting has gradually evolved over the past decades, the ESG space presents all kinds of additional concerns, including the robustness of the various data sources and processes they now deal with. One delegate told a panel that she was tired of constantly thinking about risk and wanted to start thinking about creating business value.

It makes little sense for these organizations to start integrating generative AI from the start. They must embrace this disruptive technology as it evolves through the platforms they already use. Vendors like Workiva and the management consultants and auditors working in this area are ready to go a step further.