Bank of America is using artificial intelligence (AI) to reliably predict which companies are likely to be acquired close to a year in advance — just one way the $2.2 trillion bank is employing AI in its trading business.

At a North America Fintech Connect virtual conference today, Rajesh Krishnamachari, global head of data science at Bank of America Merrill Lynch, said that “by combining macroeconomic data and idiosyncratic factors, technical factors, even market variables, you’re able to say with high probability: this company might get acquired for these reasons, and enables our bankers to target these companies better.”

Machine learning allows the bank to predict volatility by moving beyond linear regression to use statistical techniques that account for non-linear factors such as data outliers, correlations and regime changes, Krishnamachari said, adding that this helps traders better position themselves “across equities, FX, credit, mortgages and commodities.”

In a rare look at how major financial institutions are leveraging AI investments to address a variety of banking challenges, tech leaders from PNC, Bank of New York Mellon, BMO Financial and Citizens joined Krishnamachari on the panel on technology’s role in overhauling banking infrastructure, and shared how their organizations are deploying AI.

While most said they AI to enhance the customer experience, the spectrum of use cases discussed showed how varied AI deployments can be.

PNC

The $433 billion PNC uses robotics and AI “across the spectrum,” said Chris Ward, executive vice president. Robotic process automation (RPA) bots are used to fight fraud by automating the return of money to the compromised account, “crediting back to the customer much more quickly,” Ward said. The bank also uses AI to forecast spikes in mortgage transactions based on application. The Pittsburgh, Pa.-based bank also uses an initiative called “Sherlock” to analyze the products that similar customers have used to provide recommendations to relationship managers or treasury management, based on that data.

Bank of New York Mellon

Bank of New York Mellon supports a “very significant, growth market infrastructure,” working with institutions across the world, said Roman Regelman, CEO of asset servicing and head of digital at the $1.9 trillion bank. To that end, the bank uses AI to predict trade fails so humans can intervene to ensure the trade goes through, Regelman said. The bank also uses AI to parse signals from the millions of client inquiries the bank receives to decide which must be routed to human customer service.

BMO Financial Group

The Bank of Montreal uses AI to identify cyberattacks, said Ren Zhang, chief data scientist at the $799.1 billion bank. Like at PNC, the digital team also uses AI to personalize the customer experience. And on the business banking side, AI is used to help ensure the bank has the right price for posting a deposit; that information is then provided to a banker to help with the negotiation process, Zhang said. Ntural language processing is employed to analyze customer surveys for marketing, and other AI tools are used to provide analysts with a first-read analysis of financial statements.

Citizens

Citizens has focused its AI effort on improving the customer experience by offering personalized cross-sell opportunities, said Chief Information Officer Michael Ruttledge. The $176 billion bank also uses natural language processing to perform automated reading of some paper processes, streamlining customers’ ability to engage with the bank. And, the bank’s IT team also uses AI to predict any incidents or problems that might lead to outages in order to head them off before they occur.