%202.png)
Earlier this year, Georgina Sell, Principal Consultant, EMEA - Industry & Infrastructure, shared a paper ‘The Green Recovery and Beyond: How exactly are industries working to combat climate change?’, inspired by Project Drawdown’s roadmap.
Recently, Georgie spent time with Lorenzo Saa, Chief Sustainability Officer at sustainable technology company Clarity AI, to learn about the key tech drivers accelerating sustainability and yield a better understanding of AI and the opportunities associated with its ability to address climate change.
Can you tell us a bit more about Clarity AI?
Lorenzo: We provide sustainability data and insights to investors, consumers and companies to bring societal impact to markets.
Our approach is evident in our name: Clarity. The intention is to bring transparency to the traditional black box ESG data and insights offering, to ensure there is a level of traceability regarding data and the underlying methodologies, so investors can focus on driving their own sustainability investment choices.
In fact, compared to when I started my career 20+ years ago, investors today are much more competent and knowledgeable on sustainability and make decisions by looking at the underlying data rather than just relying on an external service provider’s ratings.
As for AI, which is also reflected in our name, we recognized that it could be a powerful tool for enhancing scale and speed while addressing the complexities of sustainability data. AI is not the end goal but a means to achieve it, complementing a data-driven approach rather than relying solely on analysts.
Why do you think AI could be a significant force in addressing sustainability?
The world has made some big sustainability commitments. However, we're very far from the Paris Agreement goal of 1.5 degrees today as we are currently on the trajectory of a 2.8 degrees world compared to pre-industrial times. Moreover, only 15 per cent of the SDGs achieved. Finally, we are quite behind the Global Biodiversity Framework “30 by 30” conservation target, with only 17 per cent protected area on the terrestrial side and eight per cent protected area on the maritime side.
With all this in mind, we need the right level of speed and scalability to address these complex problems hence the importance of technology and AI.
We believe any consumer, company or investor makes decisions along the ‘knowledge pyramid’. Data is at the bottom of the pyramid and as you move up there is information, knowledge, wisdom and action. We drive action by bringing AI and subject matter experts together at each step of the pyramid.
How can we use AI to help strengthen ESG data?
I think there are four elements - the first is coverage. AI can collect data on web pages or PDF documents and read free text, graphs, and images; combining humans and AI can simplify and expand this process and significantly expand data coverage.
The second element is reliability. Ensuring the quality of the data is still a huge challenge, but machine learning reliability models can perform specific checks to drastically enhance accuracy.
The third one is alternative data sets. When reported data is not available, we have different alternative data sets to leverage on. In this context, AI enables us to transform unstructured data (like satellite data or news streams) into structured data, providing deeper insights.
Finally, estimates play a crucial role. Finally, when data just isn’t there the estimation capabilities of AI are particularly strong: Estimation helps investors move ahead on climate issues which is important from a scope 3 perspective. These indirect emissions represent around 70 per cent of the average corporate value chain’s total emissions. As 42 per cent of companies report on scope 3 and usually on only one of its factors. Estimation is a critical element to address a significant part of the transition to net zero.
Where do you feel ESG data estimations and AI may not work?
That's a great question and I think we should be first and foremost encouraging the market to drive the reported data. The fact that companies are still not reporting certain issues is a challenge and I think there's a lot of initiatives to drive that change.
However, when data is not available, you need to ensure that using estimation is appropriate by assessing a few key elements:
Firstly, looking at the methodology and how critical that estimation is and secondly, evaluating the assumptions behind the estimation - are they transparent and do you fully understand them?
The third is the confidence level of that estimation. Whether it’s high, low or medium, this will help you determine if using a lower-confidence estimate might be too risky.
The final point is ensuring estimations start from a validated data set for comparison. I think sometimes there's this feeling of ‘let's not do estimations, it's not the right thing’. The reality is estimations are used a lot in financial accounting and we don't see why we shouldn’t look at them from a climate perspective.
Once these aspects are critically reviewed, then we should remind ourselves that the issues that we're trying to address are so significant that an 80/20 rule should be used to move forward. Like we do in financial accounting, estimations are not perfect, but they can still be close enough and practical to take action against them.
How does this fit with the needs of SFDR/CSRD for asset managers and corporates?
As the reporting landscape becomes more complex, with directives such as the CSRD and SFDR, AI can help simplify the process. For example, in the case of the SFDR (Sustainable Finance Disclosure Regulation), which mandates that market participants report on the sustainability impacts of their investments, Clarity AI offers a tool that allows you to automatically generate your Article 8 periodic reports using a simple questionnaire, our data and previous reports.
This is important to us as it allows investors and corporates to focus on the ‘doing’ versus the ‘reporting’.
How do you think AI can make life easier for investors, companies and consumers to help them report and progress on their own sustainability journeys?
Transition plans are critical in climate and require a lot of review to see if they are credible. We are now capable of looking at these reports through AI and assessing if the targets they set are accompanied by quantifiable measures that can achieve those targets. We recently found for example, that only 40 per cent of the higher emitting companies have these credible transition plans.
Additionally, AI assistants, like the one we’ve developed at Clarity AI, use advanced algorithms to interpret and contextualize ESG data, providing insights and trend analysis. This enables investors to “converse” with their portfolio, finding sources, trends and potential areas of improvements in a much easier and efficient format.
These are just two examples of how life can be much easier with the leverage of AI.
How can we counter risks associated with AI?
I like to think there is an ESG of AI. Governance is probably the critical starting point and a good way of looking at it is how we manage our controversy solution. It runs on discriminative AI, not generative AI, to avoid hallucinations, it removes names of companies at the beginning of the process to avoid any bias towards certain names, and any controversy deemed at a high severity level must go through one of our subject matter experts to ensure that we best manage the combination of AI and humans in the loop.
On the environmental side, there's a need to consume electricity to run the algorithms and to cool facilities, which uses a lot of water consumption. These issues need to be addressed, starting by using renewable energy, but also by designing efficient models better aligned to their use cases, and investing in further hardware efficiency innovation.
There are different ways to look at the social aspects, for example in terms of what's going to happen regarding labor. As in climate, AI will also require a just transition, with companies and governments taking their part in addressing it.
How do you believe AI can help empower consumers to accelerate the transition to a more circular economy?
We're working with ING Spain to provide over four million customers with the possibility to see their carbon footprint on their banking app based on their activity.
How you act as a consumer to ensure that your footprint gets reduced will become part of the way we think of our activities, if we have something in our pocket that informs us of our sustainable actions.
What do you foresee as the most essential skills to drive what is needed in AI?
I would argue everyone will need to have a balance of technical, analytical and soft skills, and - depending on the role - the weighting of these will change. Not everybody needs to know the details of NLP machine learning (cloud computing), but a general understanding will probably be needed by everyone.
How to use these tools will be quite critical alongside wanting to continue to learn. A great quote that I heard was that AI is not going to take your job, but someone that knows about AI will take your job.
And so being open, willing to see where the world is going and wanting to learn new things will make a difference in terms of success.
But you don't all need to be technicians – we have a huge data science team, but we also have a significant number of subject matter expertise. The ability to collaborate between data scientists and sustainability expertise is what we believe drives our success.
Clarity AI recently won the Impact Investing Platform of the Year, you were also selected as technology pioneer by the World Economic Forum and named one of the most innovative companies. Why do you think Clarity AI is so award-winning?
We are very honoured by these recognitions and most recently, we were also named a leader in Forrester's Q3 2024 ESG and analytics provider assessment. We are driven by two things: One is the desire to stay on top of innovation, which is critical in today's world. I think being able to know how to leverage tools and technology to drive scalability is what investors are particularly keen to see.
We are creating waves in the sustainability space, and we're excited about that, but we don't just offer data, methodologies or tools, we combine the three which is quite unique in the market, so we find players extremely excited to work with us.
What is your biggest call to action?
Three things spring to mind. The first is we are all a force of change and have a role to play, so it’s important to recognize this, take responsibility, and act.
Secondly, I think the world has this simplistic view of how to address climate change with black and white approaches. I think a more nuanced engagement with companies, governments and people will make the difference.
And the last thing is embracing technology. There's no way we will meet the Paris Agreement target if we don't use our intellect, creativity and innovation to drive change. A lot of the solutions are out there but they need to be scaled.
Instead of thinking of AI as the Terminator that's going to inform us, the way I look at it is that we're all Skywalkers and the Force is with us. But Skywalker had R2D2 with him.
And that's what AI is for us.
