Enhancing compliance with AI 

6 minutes01/22/2025

This article was originally published by Enhesa – a trusted Quentic partner - summarizing the discussions held during a webinar hosted at the end of 2024.

Expert insight from the Chief AI Officer at Enhesa, Director of Sustainable Finance at LSEG, and a leading GRC analyst on the future of AI in sustainability and compliance.

AI is changing the way we do business, with the potential to improve resilience through better risk management. It is increasingly being used to support strategic business initiatives, and compliance and sustainability are no exceptions. Why? The regulatory landscape is inundated with new and revised laws every single day, putting immense pressure on companies to track hundreds of changing bills in their sector and jurisdiction to maintain compliance.  

Emerging AI technologies can help increase teams’ efficiency and agility, reducing the burden of excessive time, resources, and money spent on mitigating these risks. Chief AI Officer at Enhesa, Alexander Sadovsky joined the Founder of GRC 20/20 Research, Michael Rasmussen, and Director of Sustainable Finance & Investment with the London Stock Exchange Group, Marianthe Evangelidis, to discuss the potential impact AI can have on compliance and sustainability.  

Here, we provide a summary from the Enhesa’s 2024 webinar on AI as a strategic business initiative.   
 

1. How automation in AI can enhance compliance efforts 

“Modern business is navigating chaos.”  Michael Rasmussen, lead analyst and founder of GRC 20/20 Research 

Quite simply, Michael Rasmussen explained, businesses are struggling to keep up with the mass volume of regulatory change. New and pending regulations, updated guidance, and enforcement actions are transforming the risk environment, generating geopolitical, economic, and environmental uncertainty.

On top of changing regulations, businesses have to keep up with internal changes. Are employees who move into new roles aware of sustainability policies? Are those that traverse new areas in the organization deal with new technology, and gain responsibility for new jurisdictions trained in compliance?

These challenges can trigger gaps in compliance, resulting in fines, penalties and even convictions bestowed upon companies who can’t “keep up”. The risk of non-compliance increases with each new regulation. In turn, financial and reputational damage can lead to loss of market access and opportunities.

Newly published laws can be hundreds of pages long, demanding excess time and resource use from humans — but with technology like AI, making sense of these regulations can be significantly accelerated, minimizing the risk of non-compliance. 

Chief AI Officer, Alexander Sadovsky, outlined the three upcoming technologies revolutionizing the way businesses can automate regulatory sustainability and compliance.

Content creation

Rather than employees wasting time searching for changes and new laws, AI can screen for new content automatically, finding relevant laws for business sectors and regions and changes to existing bills.   

It can be challenging for global teams to correctly interpret the volume and complexity of regulations. EHS compliance teams could miss something given so many new regulations or changes. Plus, they must correctly interpret all this information for their business, and plan and track the right steps to mitigate the risk of noncompliance.   

As Michael initially mentioned, there’s “miles of law” to sort through. Alex noted that AI can summarise whole law documents so businesses “know quickly what’s happening when”. It can also place documents into context so companies can identify relevant information more quickly. For example, AI can compare legislative documents to identify global impacts of similar laws, or extract specific information, such as the legal obligations specified throughout a 100-page PDF. Furthermore, AI is able to predict future trends based on past patterns.

Content discovery

“More data equates to more problems.”  Alexander Sadovsky, Chief AI Officer at Enhesa

AI tools enhance searchability, allowing businesses to find nuanced answers to their problems. Additionally, AI can utilize photos to achieve results. For example, EHS teams can upload a photo of an office while conducting a health and safety assessment. AI can then analyze these photos to identify potential issues, such as missing labels on appliances or safety concerns with ladders or wires.

Content enhancement 

AI, Alex noted, is helpful in relaying information that businesses might not necessarily know how to express. For example, there may be a law around carcinogenic chemicals, which auditors may not necessarily be familiar with as they don’t often work with chemicals or sustainable chemistry practices. In this instance, AI can cross-reference data, dive further into the chemistry content, and relay this information to suppliers, creating chains of knowledge across industry.

Disclosing data  

Marianthe Evangelidis, the Director of Sustainable Finance & Investment at the London Stock Exchange Group (LSEG), expanded on the enhancement capabilities of AI in disclosing data to public domains.

The main challenge facing investors and clients is that every business discloses their sustainability information differently, so it can be difficult to consolidate the right data to disseminate to regulators. Decentralized EHS compliance can create silos and gaps, leading to inconsistent, incomplete, and inaccurate data that can threaten the reliability of a company’s compliance program.

Once again, AI can be used to simplify and streamline this process, scanning for publicly disclosed information, sustainability reports, and extracting data for auditors and suppliers.

2. How to avoid the potential pitfalls 

How can companies adopting the use of AI to improve compliance efforts avoid the technological, philosophical, and legal pitfalls?

Can you truly trust a machine?

“AI is only as powerful as the data and experts behind it.”  Alexander Sadovsky

“There’s a lot of issues around how much faith you put in AI,” Alex said, as every AI system is different and requires testing to check its validity, understand its accuracy, and identify which use case it’s appropriate, or not appropriate, for. One way of managing this potential risk is through operating a stoplight system:  

• Green = information is verified and good to use
• Yellow = information should be checked
• Red = the AI isn’t at that stage of capability yet, and so shouldn’t be relied upon

“AI is a search engine, not a lawyer”, and isn’t designed to replace human intelligence, but rather augment it, Alex said. 

“Garbage in, garbage out”

Making sure you have the correct data going into your AI model is an important step in reducing a major pitfall in using these tools, Marianthe noted. The data needs to be accountable, trustworthy, reliable, and usable to ensure AI output is efficient.

Compliance data is often spread all over an organization, requiring a single platform to streamline findings. Opting for a generic AI model can be both inefficient in terms of energy usage and inaccurate with unreliable sources. Enhesa’s specific content library uniquely places us to output high quality results, compared to a generic AI model which risks being incorrect. A bespoke model, on the other hand, “can be powerful,” said Michael, and “doesn’t replace the subject matter expert, but extends [their knowledge].”

3. How AI can transform sustainability and compliance outcomes 

Governance, Risk, and Compliance (GRC) on values

“GRC by definition is the capability to reliably achieve objectives, address uncertainty with risk management, and act with integrity.” Michael Rasmussen  

Michael discusses the four areas of value to be found in leveraging AI in business cases:

1. Efficiency

“Time saved; money saved,” Michael summarized, identifying that a lot of staff time could be spent in unproductive work that’s not contributing to the company’s end goals. A machine, like AI, can work through documents quickly, saving time, energy, resource use, and money in the long-term, and allowing employees to dedicate their time to other avenues.

2. Effectiveness

AI can help employees reduce the risk of data slipping through the cracks, acting as a “copilot” in researching and presenting information to amplify the existing findings.

3. Resilience

“How will this automation help me with compliance issues?”  

An AI solution can discover new areas and gaps in compliance, and is capable of identifying small issues before they grow into larger problems for the company. In this way, AI can allow companies to work proactively, rather than reactively.

4. Agility

Also referred to as “horizon scanning”, AI can help businesses navigate their current and forecasted regulatory path with more agility, answering questions of:  

• “Are our eyes on the road in front of us?”  
• “Where are we heading?”  
• “What’s our strategy?” 

LSEG on data  

At the London Stock Exchange Group, Marianthe noted that AI is “a copilot”, used internally to track changing regulations. It’s crucial to “understand synergies between regulations, so you don’t keep [interpreting] the same data again and again.” Externally, AI helps the company mitigate financial risk. By allowing a deep-dive into the copious data, LSEG can interrogate it for meaningful answers to present to the board or clients. 

What could AI do for your business? 

A potential catalyst for supercharging business operations, AI can transform approaches to compliance and sustainability, making business objectives more attainable in a faster timeframe. It’s important for industry leaders to learn how AI can complement compliance initiatives to stay ahead in a rapidly changing landscape.