About a-team Marketing Services
The knowledge platform for the financial technology industry
The knowledge platform for the financial technology industry

A-Team Insight Blogs

EthicsAnswer Creates ESG-Trained AI to Ease Disclosure Pain

Subscribe to our newsletter

A SaaS platform is using generative artificial intelligence (GenAI) to significantly reduce the burden of sustainability disclosure requests from investors and other stakeholders faced by chief sustainability officers (CSOs).

Welsh company EthicsAnswer says its GenAI product will substantially reduce the time it takes for CSOs to complete ESG surveys and will improve accuracy. The tool is backed by a large-language model (LLM) that is trained on more than 46,000 publicly disclosed sustainability reports and publications. This alongside the fact that all answers are provided with evidence to maintain a human in the loop check ensures that answers are accurate and not hallucinated, said AI Ethicist Tess Buckley.

“We are using GenAI to dramatically reduce the burden of sustainability disclosures that have been – and are – overwhelming chief sustainability officers,” Buckley told ESG Insight. “Our ESG-specific LLM and embedding model ensure that the generated responses are  domain specific and have a higher accuracy rating than sustainability professionals might get from  freely available general  AI solutions.”

Overwhelmed Officials

EthicsAnswer was born from the development of EthicsGrade, ESG benchmarking agency formed in 2020 that assesses the corporate digital responsibility (CDR) of companies. EthicsGrade provided the first data set on digital ethics to investors. When gathering information for its own databases, EthicsGrade found a lack of willingness to engage in data collection from sustainability professionals.

The growth of the ESG space has meant that more investors are requesting sustainability data from their portfolio companies as part of their own due diligence processes and to comply with increasing obligations placed upon them by regulators. These often take the form of paper or digital questionnaires and surveys that can be very detailed, with as many as 50 questions. .

“We found that, for CSOs, the time it takes to reply to each  disclosure request means it is virtually impossible to respond to all of them,” Buckley said. “Imagine it takes two days to complete one disclosure request, during which time sustainability professionals can get nothing else done. When they return to send of this request they are met with a new one in their inbox.”

EthicsAnswer found during its engagements with CSOs that many were devoting lots of their and their team’s hours to filling in ESG surveys. Some had even hired employees dedicated only to completing ESG disclosures. The process of data collection across the building even described by some as a task of “running around with a clipboard”.

“It’s extremely resource-consuming and it means sustainability professionals are spending their time disclosing on sustainability rather than strengthening company ESG strategy,” Buckley said.

Evidence Backed

After considering potential time- and resource-saving solutions – including better ways to write their surveys – EthicsGrade concluded that the only way to streamline the process and achieve their mission of answering any ESG question on any company for any stakeholder was to help companies complete their disclosure requests quicker. Also, EthicsGrade realised that it already had the technology necessary to auto-populate answers. It had developed this capability to enhance the user experience for individuals engaging with its CDR surveys.

Available as a SaaS platform, the EthicsAnswer solution speeds the process by enabling companies to ask the GenAI platform  individual questions from ESG surveys or upload the entire survey to be completed. For individual questions the platform will then examine the company’s own publicly disclosed sustainability reports and generate an evidenced answer within seconds. For complete surveys it takes slightly longer to complete the full survey.

Because surveys distributed by investors often ask very similar questions, EthicsAnswer has made it possible for users to save the generated responses for future use. Additionally, it facilitates the uploading and retrieval of evidence to support those answers. Responses are offered as simple yes-no answers and as evidence-backed longer answers.

Buckley stresses that the platform can only provide answers based on previously publicly disclosed data or private data uploaded by users.

“In this way, EthicsAnswer doesn’t make anything up,” she said. “We’re not looking to display facts in the most pleasing way or answer the question in the best way; we’re simply looking for factual answers to the question.”

Critical Training

The importance of training the GenAI on ESG reports is essential for several reasons; it ensures that the technology only scours and retrieves relevant data, and that the correct meaning of general words that have specific applications to sustainability are searched.

As an example, the word “scope” has many uses but within the ESG space it refers mostly to the three forms of emissions reporting laid down by The GHG Protocol. Being able to refine the meaning of language in this way helps to improve the technology’s ability to retrieve the correct evidence.

The benefit of this became evident in EthicsAnswer’s interviews with sustainability officers who had sought to save time in their reporting processes by using the public GenAI platform ChatGPT. They very quickly ran into trouble.

“A key problem with sustainable professionals using ChatGPT, apart from the leaking of confidential data, was that its LLM is trained on the internet, it’s not specifically built for the purpose of ESG disclosures and sustainability,” she said. “And there are the hallucinations!”

By generating evidence-supported responses in EthicsAnswer “there is no greenwashing and no fancy sustainable storytelling – it literally is cutting to the chase in the evidence that’s already been produced from the company, providing structure to unstructured data”.

Subscribe to our newsletter

Related content


Recorded Webinar: ESG standards and taxonomies – A progress report

The development of ESG reporting and disclosure standards and taxonomies has progressed rapidly over the past few years, yet there is still no certainty on whether there will ever be measures that can be applied on a global basis. The EU leads the way in terms of ESG investment and regulation, ahead of Asia-Pacific and...


Banks are Tackling Unique Challenges with ESG Data, Says CGI

Non-retail banks are wrestling with unique data requirements and challenges as they strive to build their ESG capabilities. A combination of factors including weak reporting standards and legacy technology is slowing wholesale lenders’ efforts to gather and use data needed to green their loan books and operations, according to IT and business consulting giant CGI....


AI in Capital Markets Summit London

The AI in Capital Markets Summit will explore current and emerging trends in AI, the potential of Generative AI and LLMs and how AI can be applied for efficiencies and business value across a number of use cases, in the front and back office of financial institutions. The agenda will explore the risks and challenges of adopting AI and the foundational technologies and data management capabilities that underpin successful deployment.


Practical Applications of the Global LEI – Client On-Boarding and Beyond

The time for talking is over. The time for action is now. A bit melodramatic, perhaps, but given last month’s official launch of the global legal entity identifier (LEI) standard, practitioners are rolling up their sleeves and getting on with figuring out how to incorporate the new identifier into their customer and entity data infrastructures....