Why sustainability data sharing needs to start with a trust framework

George Lawton Profile picture for user George Lawton February 7, 2024
Gavin Starks, CEO of IcebreakerOne, who co-chaired the birth of the Open Banking Standard, wants to do the same for sustainability data. He believes industry-wide data sharing needs to start with a trust framework rather than fancy tech or a cloud monopoly. What he has to say has important implications for accelerating sustainability efforts and all industry-wide data-sharing collaboration.


Streamlined data sharing is one of the biggest challenges when it comes to decelerating the pace of climate change. Many technology proposals, data-sharing frameworks, and carbon credit marketplaces are stepping in to fill this void. Not to mention billions of dollars in government incentives to help nations build green economies. However, these efforts are constrained by the lack of transparency into which incentives and investments will have the biggest impact. 

The fundamental problem is the lack of trust and engagement among the executives, lawyers, risk managers, and regulators, who ultimately need to sign off on any efforts to increase transparency to gain traction. Gavin Starks, founder and CEO of IcebreakerOne, who also co-chaired the birth of the Open Banking Standard, now hopes to apply the lessons learned in that effort to Perseus, an effort to improve sustainability data sharing for small to medium businesses. The project is starting with the low-hanging fruit of already digitized electricity data.

I attended his Open Data Industry Panel session that was part of the State of Open Con 2024 in London and sat down with him for a deep dive into how they are building traction for the new project, what he learned from the Open Banking Standard, and what it means for data sharing more broadly.

One big takeaway I will elaborate on below is the critical importance of hashing out a trust framework at an executive level before trying to implement the latest whizz-bang technology, cloud service or digital twin framework.  Second is trust frameworks' critical role in accelerating responsible and effective AI adoption. 

Gavin says the finance industry spent something north of £40 million pounds to set up the open banking infrastructure, which has already generated about £6 billion for the economy and has grown into 80 countries. Now, he wants to do the same thing for energy and water. He explains:

We've been coming into this saying, ‘Well, can we do that again for utilities, energy and water?’ It's been fascinating saying, ‘Well, we've done it over here.’ And people say, ‘Yeah, but energy is not the same as banking.’ And it's just data. We're entering into this massive de-centralized energy future where everything's going to be a generator. And storage and consumption of energy is a hugely federated network. 

What architecture do you think is going to work for this? We could spend 25 years doing what we did with the web and letting individual monopolies exist and harvest everything and end up with another monolithic set of unicorns leading the whole thing, and it will have a scrappy kind of thing underneath it. Or you could actually create a proper open market, where you probably will have some unicorns in there as well. But really, can we make this for everyone? And because it's, it is actually for everyone, because everybody needs electricity?

It's important to dis-ambiguate open data from open access. Open data is about openly sharing data across parties. However, this is not sensible when it comes to sensitive financial, business, or infrastructure data. Most of what Starks elaborates on here pertains to open access. However, there is an important role for openly sharing data about standards, agreement frameworks, and trustworthy APIs to guide teams in other countries or domains to jumpstart their efforts to implement similar trusted data-sharing schemes. He describes his journey from open banking to open energy:

So because I co-chaired open banking, I kind of watched this whole thing play out. It was like, ‘Okay, so there is a need here for a neutral intermediary.’ That's not government. That's not industry. That needs to take the mandate from both to help everybody come together and bash heads together to make stuff happen.

That was the foundation of IcebreakerOne. We set up as a non-profit to be that sort of intermediary because it's a necessary thing that has to exist in the marketplace. So, the combination of that and prior to open banking, I was running the Open Data Institute. Obviously, we brought in all of our kind of open principles to that process. And with banking, just to kind of bottom that story out. When we published the open banking standards, I mandated that it was openly licensed. So, all the documentation is Creative Commons Attribution. All of the code is MIT. So it's a completely open basis to it. Because we did that, it's been copied in 80 countries.

Start with engagement

The essential starting point for implementing a data-sharing framework needs to start with a compelling vision of how the new framework will solve existing problems, provide new value, and manage risks. In the case of open banking, the UK banking industry had the ‘luxury’ of persistent regulators threatening action unless urgent progress was made. In the case of Perseus, Starks started his presentation with a short Hollywood-quality video that paints a vivid picture of what the new framework means for sustainability, small businesses, and risk managers. Stark says:

That takes quite a lot of effort, and my colleagues at ODI know very well the vast amount of time it takes to get anybody to understand what data is.

In the case of carbon reporting data, banks, asset managers, insurance companies, and landowners face mandatory reporting requirements on the carbon impact of loans and investments. This is a tremendous burden for SMBs being asked to fill in details not only about their own operations but also about their third-party suppliers, representing a massive federated data challenge. Today, people just estimate, and as we start to add up multiple levels of estimation across supply chains, the margin of error compounds.

Getting the reporting right has tremendous importance for better accounting, audits, insight, capital allocation, and brand awareness. The goal of Perseus is to create a trustworthy auditable framework that can directly correlate energy generation in half-hour increments with energy usage data pulled from smart meters to create an auditable trail. 

Businesses should just be able to go in and tick a few boxes on their energy provider and accounting system like Sage or QuickBooks and have the information flow in automatically. Also, they need to be able to trust that it is not shared with competitors or in a way that might adversely affect loan approvals. Regulators also want to ensure that adversarial nations or bad actors don’t have access to data about increased power usage near military sites.  

Better information could also unlock tens of billions of pounds in loans that banks want to make to SMBs on the basis of their sustainability. It could also make it easier for new startups that can cost-effectively help SMBs improve energy efficiency by identifying opportunities for improvement.

Building a trustworthy framework

It’s also important to note that the whole trusted framework is architected without needing a blockchain or cloud service monopoly in the middle. Starks explains:

One of the critical design features is there is no central database. Also, there's no blockchain. You don't need a blockchain...The principle here is that the trust framework stores the credentials of who's allowed to play in the system. To run an analogy with banking, if I want to share my current account data at Barclays to a Sage Accounting package, both of those parties say, 'Do you consent to this data transfer?’ Once I click consent, the data flows directly from the bank to accounting. So exactly the same architecture is used here where the data flows directly from wherever the energy data is being stored, and that happens in many places, into the accounting, 

There are two sets of controls. One is legal. The trust framework has everyone sign a contract that says, ‘We will do this, and we will not do this.’  So, if you break that, you get into trouble. On the technical side, if you revoke the keys from the API, you can switch the tap off. So, there's a technical level protection and a legal level protection and those things combined at the heart of the trustworthy framework.”  

My take

Having spent three decades in Silicon Valley, I found myself buying into the notion that better technology on its own might help us solve our many sustainability problems and maybe the regulators are getting in the way of progress. After sitting with Starks, I am starting to question that assumption. Technology will only get us so far unless we begin with a solid foundation of trust. 

Increasingly, a regulatory imperative can help when applied in the right way to focus everyone’s attention on solving the most pressing problems. This can only happen when this process includes industry, trade groups, third-party advocates, and regulators. In this case, the regulators can provide a redress mechanism when parties break their agreements. 

A Wall Street Journal article making the rounds last week argued that Europe was going to miss out on the race to new opportunities in AI and 5G owing to its strict new AI regulations. It failed to note that European phone service costs about a quarter as in the US for the same data usage. Oh, and you can make a bank transfer directly from one bank to another instantly for free; no third-party app or paper check is required. It can take three days in the US, not to mention wire fees. 

In the recent Accenture Technology Vision report, they observed that a top priority for businesses in getting ready for AI lies in sorting out their data infrastructure. This is hard to do, even internally, much less across companies. Starks, who has worked on several data-sharing projects, observed that most of these, particularly when they start with the technology stack, never move past a proof of concept. Rather, success begins with building a framework for trust and then building the technology on that shared agreement. 

A grey colored placeholder image