The term "artificial intelligence" was first coined over 70 years ago, by computer scientist John McCarthy at the famous Dartmouth conference in 1956, says Pierre Olivier-Haye, CTO, Iceberg Data Lab.
Today, the term is used every day across business, the media, in government and in our everyday conversations. In fact, it was reported that "AI" was mentioned more than 200 times on 2023 earnings calls by Meta, Microsoft and Alphabet.
This, coupled with the prediction that by 2040, between 1.1 million and 1.6 million of UK businesses could have adopted AI technologies shows that AI's influence is unmatched.
Generative AI sits under the umbrella of artificial intelligence, meaning models that use neural networks to identify the patterns and structures within existing data, to generate new and original content. You do not have to look back as far as the 1950s to see how Generative AI has developed and transformed the world we live in.
From the introduction of Google's BERT project in 2018 to the much faster-paced ChatGPT by OpenAI which took the world by storm in 2022 Generative AI continues to rapidly evolve.
For financial institutions in particular, Generative AI is an invaluable tool - yet its full potential is largely unknown and untapped. There are several existing use cases of Generative AI in finance today - such as for financial modelling, risk analytics and investment - decisions however, many don't yet utilise the unrealized contribution this technology can offer to ESG investors, which in turn can support the industry in speeding up the climate transition.
Elevating ESG investing with Generative AI
Combating climate change and biodiversity loss is a global responsibility, but financial institutions play a critical role in our path to net zero. To address and mitigate our environmental impact on the planet, players across finance need to allocate capital to the right places, promote less carbon-intensive activities, and support companies on their path towards sustainability and carbon neutrality.
Nevertheless, understanding our portfolio's impact on climate change remains notoriously difficult and complex. One of the most difficult measures is biodiversity loss - a company's impact on the natural world is much harder to measure than climate change. Impact on biodiversity loss is measured across land use, water pollution and air pollution, with the parameters for measurement vastly differing across each.
This is where Generative AI can step in. By using data-driven solutions, financial institutions can drive capital towards environmentally sustainable initiatives. The ability of this technology to gather and examine multiple datasets offers a deep understanding of any company's operations, processes, and position in the value chain in relation to global climate impact.
For example, at Iceberg Data Lab we have capitalised on advancements in Generative AI to build an ESG AI assistant that handles complex and specific ESG issues. This assistant (who we call ‘Barbatus') is trained to extract information from over 2,300 products within our datasets and analyses information across all scopes, sectors and countries - helping analysts to easily understand complex ESG data on any portfolio company.
The benefits of generative AI to the financial services sectors are clear to see. Indeed, companies like Magnifi use the infamous generative AI product, ChatGPT alongside computer programs to provide personalized, data-driven investment advice. Not only can Generative AI enable products such as Magnifi's to answer investor questions in human-like conversations, but it can also monitor individual portfolios, guiding users through market-moving events like rate hikes and earnings reports.
Abundance of opportunities, vigilance of risks
Whilst Generative AI is a brilliant tool for guiding sustainable investing, there are a number of risks to be aware of.
Transparency of data is an obvious one. How do we really know if we're being told the truth? Promoting responsible AI practice is key. At Iceberg Data Lab, our ‘Barbatus' assistant generates real-time, text-based and fully sourced explanations in response to questions regarding the ESG data of portfolio companies.
We believe that addressing these accountability issues is the responsibility of those developing and deploying these products to ensure outputs for clients are unbiased, sourced, and accurate.
Data privacy and data bias are widely discussed issues of Generative AI, moving companies such as Apple to forbid the use of ChatGPT.
However, the finance sector is well-equipped to meet these challenges, with due process already in place to mitigate data breaches and leaks and should be committed to upgrading processes as technology continues to rapidly advance. For example, at Iceberg Data Lab, we have in-built privacy and data processes into our products and takes steps such as only inputting sections of files into our Generative AI models to combat against privacy concerns.
AI replacing jobs is another risk levelled against the technology. Like any technology, Generative AI serves a purpose in supporting us in our jobs to carry out the time-consuming, cumbersome and repetitive tasks with more efficiency, so that we can focus on more abstract, strategic and creative ones. Perhaps the best analogy here is how the internet was once a new, unfamiliar system, which is now revolutionary - at the click of a button anyone can access a broad of research, thus transforming the way in which we work. AI can do the same.
I am optimistic about the opportunities Generative AI can bring to ESG investing and in turn how the decisions it drives could positively impact our planet. Every bank, asset manager and brokerage must examine the impact of every company they invest in as well as its value chain to drive capital allocation away from polluting industries and companies and towards a greener future.
By Pierre Olivier-Haye, CTO, Iceberg Data Lab
