How AI Can Help Meet the ESG Challenge

10 March 2022 | Luke Taylor

Originally published by ThomsonReuters © ThomsonReuters

Demonstrating environmental, social and governance (ESG) credentials is a mammoth task for organisations, which must be able to identify and assemble trusted data across their entire supply chain, ensuring transparency in all operations and jurisdictions.

In the financial services sector, non-financial disclosure obligations have been rapidly increasing, with a raft of legislative updates introduced across England and Wales and the EU since 2017. Recent new reporting requirements include the Non-Financial Reporting Directive (2014/95) in 2018, which requires public interest entities with 500 or more employees to disclose their policies and impact on matters including environmental, social, employee and human rights concerns, including diversity, in their annual report.

Most recently, the Sustainability-Related Financial Disclosures Regulation (2019/2088) (SFDR) came into force last year, requiring financial market participants to publish their policies on integrating sustainability risks in investment decisions.

Complex reporting landscape, operating structures

In the UK, the regulatory environment is overseen by a patchwork of authorities, including the Financial Reporting Council, the Financial Conduct Authority (FCA), the government and the European Convention on Human Rights, creating a complex reporting landscape

for firms. Companies may also be signatories to voluntary commitments, such as the UN's Sustainable Development Goals and the Principles for Responsible Investment.

The challenge is made still harder by complex corporate operating structures. Legally relevant data is often siloed; whether that be across various cloud storage environments, different computers due to bring-your-own-device policies and employees working from home, or even in the minds of employees following personnel changes.

The retrieval of enterprise data is already no simple task due to the sheer volume of data to search through. There are challenges at each stage of the drive to maintain effective ESG reporting, but with new technology such as artificial intelligence (AI) at organisations' disposal, difficulty is no longer an excuse.

The need for AI has never been greater: while rudimentary tools and techniques may help organisations to map and understand their contractual landscape for ESG issues, simply getting the "lay of the land" is no longer enough. Companies and other complex organisations need AI technology that enables them to implement their policies consistently and ensure they walk the talk when it comes to giving their commitments and policies legal force in their dealings with customers and suppliers.

ESG will continue to grow in importance in the next few years, and AI technology can help to ensure businesses continue to be aligned as companies scale, perhaps employing a wider group of people and developing operations across different territories.

For example, AI technology should allow companies to track climate risk, use of raw materials, modern slavery commitments and a growing number of other issues not only in their existing contracts but also for incoming contracts. Companies cannot afford to get this wrong, with accusations of greenwashing lurking for those who fake or exaggerate their eco credentials.

How AI can help organisations meet their ESG targets

Let us imagine a financial services company based in the UK needs to review and understand its entire contractual landscape, to ascertain its ESG position and exposure to any changes in legislation. This company also wants to include its subsidiaries across other European countries in its review.

Historically, manual contract review was the only answer to this, resulting in a significant drain on resource. Using AI would enable this company to improve its overall approach to evaluating its ESG standing in three main ways:

Providing an immediate understanding of the company's existing ESG position

For example, AI could instantly flag all anti-bribery clauses present in the contracts, as well as highlighting those where such provisions are not included. AI could also highlight out-of-the-box all governing laws across contracts, allowing the company to gauge its operations and exposure to any volatile regions or countries that have yet to sign up to the pledges made at COP26.

Helping organisations understand their documents in a more intelligent way

At present, many companies rely on manual searching or more basic technologies to help them find answers within their documents. The former involves a vast amount of time and expense, while the latter can put organisations at risk of missing something.

By contrast, AI can understand concepts within documents in much the same way that a human brain does, forming links between ideas and ensuring lawyers gain the most thorough understanding. If, for example, this company were searching for mentions of "coal" within their documents, as part of assessing climate risk in their investment portfolio, AI could recognise that words such as "emissions" or "mines" would likely also be relevant to the search, and thus should be included.

Anticipating future changes in standards or regulation by adapting as social mores or laws shift

Modern AI is so sophisticated that it can be updated simply by being shown an example of how a concept should look. AI will then refine its understanding in the light of this. For example, should financial levies be imposed on non-renewable energy sources,

AI simply needs to be shown one instance of how clauses mentioning "dirty" energy should be phrased when negotiating deals with suppliers.

Following exposure to this one example, it will then be able to flag instantly every other document containing this clause, as well as any instances of non-compliance in all subsequent incoming documents.

Wider application

AI that understands contracts is not just helpful for lawyers. Procurement functions, for example, would benefit from AI's ability to surface standardised clauses regarding ESG-related issues during their negotiations with suppliers. Human resources would benefit from AI technology which could ensure effective contract harmonisation, for instance, by flagging inconsistencies between contracts across regions or highlighting any differences in the contractual terms offered to male and female executives.

ESG reviews amount to such vast exercises that point solutions simply do not work for them. The landscape is evolving too rapidly across different requirements: compliance, new legislation, soft law, or voluntary commitments that have been made each have different effects on contractual arrangements with suppliers, customers and the rest of a business' universe. Understanding those impacts is one thing; turning them into a competitive advantage is another.

ESG now touches all legal and regulatory aspects of an organisation. Agile AI can remove many barriers to a fast, flexible and comprehensive review process by providing a system that can read and form a conceptual understanding of documents and data.

Using AI to perform contract harmonisation exercises in this way will make these reviews — and overall compliance — easier in the future by adapting to a new ESG landscape that is changing all the time, putting organisations in the best position to succeed.