KeyBank is continuing its AI and gen AI development pipeline after seeing positive effects on its operations. 

“We have roughly about 40 proofs of concept [POCs] across KeyBank that we are evaluating right now,” Ken Gavrity, head of commercial banking, told Bank Automation News. “You are going to see it across our business as those start to roll out over the next quarter or two.”

The gen AI POCs across the bank include: 

• Aiding customer service representatives; 

• Anti-money laundering screening; 

• Coding; 

• Content creation; and 

• Commercial underwriting. 

Realizing returns 

Gen AI is saving time at the $185 billion bank, Gavrity said. 

For example, manual AML solutions took several hours to conduct an investigation and properly screen transactions and entities, he said. With gen AI, that screening can take just 15 to 20 minutes. 

Additionally, underwriting has the opportunity to be streamlined, making decisioning faster and more accurate, Gavrity said. 

By using AI and gen AI tools, KeyBank can aggregate and synthesize information on public companies quickly, Gavrity said.  

AI at Key 

KeyBank’s strategy has been to implement generative AI to streamline internal tasks before eventually using it in consumer-facing channels, Robbi Armstrong, vice president and group product manager of conversational AI at KeyBank, previously told BAN. 

The bank has deployed AI for documentation, compliance, fraud detection, marketing and internal research, Armstrong said. 

The Cleveland-based bank reported noninterest expense of $1.2 billion, up 7% year over year, driven by higher tech and personnel expenses, according to the bank’s second-quarter earnings report published on July 22.