Many years ago, one couldn’t be blamed for a car accident at night. After all, there were no headlights or streetlights to see what was coming, no traffic lights at intersections and no weather apps to warn us of bad road conditions. The impact of a traffic accident was expected to be bad because there were no seatbelts, airbags or apps to call for help.
As advancements in vehicle and traffic safety have evolved, we are more accountable for anticipating what could go wrong and for using available tools to prevent it.
We wear seatbelts, use airbags and rely on signals to warn us about what’s ahead. We have weather apps, weather sensing windshield wipers and dynamic drive technology for water, snow, sand and rock (and are expected to slow down and adjust to changing conditions).
The financial services industry has evolved from paper ledger books, back-of-the-envelope calculations and the physical exchange of funds to fast-moving global markets that use sophisticated technologies and move money at lightning speed over a spiderweb of digital pathways.
As with the transportation industry, as advancements in safety measures across financial services have evolved, we are more accountable for anticipating what could go wrong and using tools to prevent it.
As an industry we’re safer than we ever have been, yet the pace and interconnectedness of markets, payments and infrastructure, coupled with a dynamic geopolitical environment, make it difficult to see what’s coming and contain an event once it has started.
An input error can cause a flash crash. News on the wire can swing markets before humans see it coming. Bad actors can use our own technology to commit fraud and disrupt markets.
As long as people walk the earth and use financial services, they will make mistakes and use judgment to prioritize risk and return. As with traffic safety, the only way to get to zero accidents is to stop driving. That’s not the goal. We need vehicles and we need financial services.
Using artificial intelligence, we can optimize the upside while managing the downside. AI can be used to predict and help prevent operational losses, drive efficiency and understand when processes are likely to fail so we can target investment where we’ll have the greatest financial and operational return.
For example, BMO created an AI-based Operational Loss Intelligence tool (OLI) to predict and help prevent operational losses. To illustrate, before markets opened in Europe, OLI predicted a 62% chance of an event in European equities on May 2, the day of a fat-finger error that caused a flash crash. With this advanced warning, people can ready themselves for a higher-risk day by slowing their input, clearing their queue of transactions and double-confirming communications.
We are also using meta-data from processes to determine when those processes, or elements of them, are likely to fail. With this information, we have intelligence to drive decision-making. For example, during annual planning processes, executives vie for finite resources required for business growth. “We need $2 million for straight-through processing,” “We need $4 million or this process can’t handle the growth we’ve planned.” Wouldn’t it be better to prioritize spend based on data?
To combat fraud, AI can combine internal and external data, find relationships among data points and then alert clients when they are facing a higher risk of fraud, prompting them to take action like confirming a beneficiary account number before initiating a recurring transaction.
Instead of incremental hires to handle volume, we’re using AI to create capacity so that we can redeploy people to activities that must be done by a human. Not only is this good for efficiency, it is also good for people who would like to participate in an AI implementation and have a desire to grow their career in a new area.
Can AI help to save us from ourselves? It is a powerful technological advancement that can help when used effectively. If you’re still not sure, the next time you drive at night, turn off your lights, remove your seatbelt, turn off your airbag, unplug all your sensors and only use your mirrors.
