The rise of advanced artificial intelligence is bringing many potential benefits to the financial industry, such as an improved ability to analyze documents, generate code and communicate with customers. But it also comes with an unintended consequence: increased carbon dioxide emissions. Some tech providers say they have answers, including the ability to shift work to carbon-neutral cloud providers and to colocation facilities in areas with clean energy.
Banks, like other companies, consume massive quantities of energy, largely through their use of technology.
"IT accounts for somewhere between 2% and 4% of global emissions right now, which is about the same as the airline industry," said Chris Noble, CEO of Cirrus Nexus. "It's a lot of carbon that's being produced." As more companies use AI, they need to access ever-growing pools of data, which requires a lot of energy.
"And more and more companies, especially in the financial sector, want to use AI for everything from risk assessment, fraud, to just basic analytics," Noble said. "If you run AI against it, you really have a much deeper view of what's going on. So that huge amount of data that's been collected for 20 years is being processed instead of a couple times a day, every second of the day."
But some banks and their vendors are trying to be more responsible and find ways to curb their use of power.
"What I have seen with banks regarding sustainability, is that they see it as an opportunity," said Falk Rieker, global vice president and head of banking at SAP. "Banks have an important role to play in the economy and in society. Therefore a lot of the institutions I talk to see it as an obligation to do something around sustainability." European banks tend to be far more concerned about carbon emissions than U.S. banks.
Climate First Bank in St. Petersburg, Florida, is on a mission to be green. It recently announced that in 2022, it offset 100% of its carbon emissions and gave $71,000 to environmental nonprofits.
Some say banks' ongoing efforts to shift to cloud computing may help. Backbase, a provider of customer-facing software for banks, is trying to become carbon neutral and about 65% of its effort is based on migrating to Microsoft Azure, which is carbon neutral,
"I think Microsoft is really the gold standard, compared to Google or AWS," said Backbase CEO Jouk Pleiter.
If a bank moved all of its computing cycles over to Microsoft Azure, it could massively reduce its carbon emissions, according to Pleiter.
"We are here as a visitor," Pleiter said of his commitment to lowering energy consumption at his company. "This is the best thing you can do next to education for your children."
Another approach is to lower the power consumption of a bank's data centers by moving workloads in real-time based on time of day, geographic location and power sources.
Clean energy sources like solar and wind power are not steady, the way natural gas and coal are, Noble pointed out.
"The sun isn't always shining, the wind isn't always blowing," he said.
Cirrus Nexus helps companies identify where and when to shift workloads to clean energy. It could be that the best time to run heavy data workloads is at night when there's low demand and there's a lot of wind, or in the morning when the sun is coming out.
"If I can run my data processing, whether that's AI or whatever it is, for my customers on the West Coast during the day and run that operation in the EU at night when there's less demand, I'm generally using cleaner power," Noble said. When the wind is blowing in Ireland, for instance, it is a good source of clean energy.
"Ireland went from one of the more carbon-intense regions to one of the cleaner regions based on that," Noble said. "In fact, they even became an exporter of power for a little while because the wind was blowing so heavily. So what companies that run a lot of data constantly need to start looking at is where they're running it and when they're running it."
In the U.S., the Northwest is generally the region with the cleanest power, Noble said, because there are several hydropower plants that produce clean energy.
Cirrus Nexus software helps companies determine where to run computing workloads, what services to use, in what regions, when to run them and when to move them around. The software uses predictive analytics to analyze weather patterns.
"If we see a windstorm coming through Ireland, we know that it's going to be cleaner there," Noble said. "We also know that the wind moves on, and then it's in Scotland, or whatever. So we know moving the workload from Ireland to London might be the next best place to run it and follow that weather." (The bank would need to have its own servers in these regions or use colocation or cloud providers in all these places for this to work.)
Large financial institutions have "meaningful but manageable" loan volumes to industries that are most likely to see declines in output and profit after climate-friendly regulations take hold, new research shows.
The software learns how to shift workloads based on a company's usage patterns. For instance, for banking apps in the U.S., peak usage typically occurs between 4:30 and 7:00 p.m., as people check their balances multiple times. For each customer, only a bit of power is used on the phone when they check their credit card balance. But behind the scenes, that tap is hitting the bank's servers, the credit card issuer's servers and the clearing house's servers; it's entered in a fraud detection system and other banks in a network may be notified.
"It's a tremendous amount of carbon that gets spent" through such simple transactions, Noble pointed out.
Another technology approach is to simply monitor carbon use more carefully and use that information to drive responsible decisions.
SAP is working with a large bank in the Asia-Pacific to report the climate emissions impact of all its travel, procurement and real estate. SAP's software can also analyze the carbon footprint of a bank's loan portfolio or of individual branches.
Such analyses are bound to produce unwelcome truths about customers or partners that have large carbon footprints.
"The insight into the carbon footprint of suppliers and eco-friendly suppliers will play a role," Rieker said. "Because companies will say, if I use that supplier, I get a better sustainability index, and with a better sustainability index, I get a better loan rate, I get a cheap loan. And the same is true of banks at the end of the day."
